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

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

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

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

#rockbox log for 2007-11-10

00:00:05 Join criznach [0] (
00:00:11przemhbbarrywardell: I do reconfigure and remake
00:00:22barrywardelli mean on the h10 itself
00:00:42barrywardellit's a long shot, but maybe an old file in the .rockbox dir is causing problems
00:01:07 Quit funglist (Read error: 110 (Connection timed out))
00:01:12przemhbbarrywardell: I do synchronise based on file contents comparison
00:01:34n1sprzemhb: wild guess: but the root menu probably calls the function that checks if an fm radio si present to decide if it's going to display the menu item, might be worth checking out...
00:04:31 Quit theclaw ("n8*")
00:10:58 Join mirak [0] (
00:11:15 Quit Domonoky ("Trillian (")
00:11:25 Join Shoot [0] (
00:11:31ShootIs anyone there?
00:11:50 Quit kkurbjun ("leaving")
00:12:44ShootI have a small question
00:12:49ShootDoes the Rockboy FPS lag?
00:13:36barrywardellprzemhb: your fmradio_i2c_getack is different than the h100/h300 one. I wonder if that could be a problem, especially the while loop
00:14:01przemhbstrange: this time it get stucked at cpu_boost(false) in init()...
00:14:07ShootAlso; how do I place .gbc files into my e200?
00:14:11ShootIt doesn't let me
00:14:13 Quit alienbiker99 ("( :: NoNameScript 4.02 :: )")
00:14:25ShootSays something like "Only pictures, videos and .mp3 can be placed in this drive"
00:14:45Nico_PShoot: your sansa has to be UMS
00:15:04ShootUSB mode MSC?
00:15:06Nico_Pand rockboy may be slow on some targets but honestly I don't know
00:15:11ShootAlright thank you so much
00:15:18ShootOk; I'll try it out
00:15:27ShootIf it'll help you guys, I'll give you a response on the FPS lag
00:15:39ShootNow... just have to find some nice GBC games >:D
00:15:55barrywardellprzemhb: I'll try it on my own H10 later/tomorrow. maybe I'll spot something then
00:17:08 Quit lee-qid ("aufwiederbyebientotsayonara")
00:17:43 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
00:17:50przemhbbarrwardell: ok; now it stops at cpu_boost(false); line in init()
00:18:14przemhbbarrywardell: exactly the line
00:18:25 Join alienbiker99 [0] (
00:18:29 Quit pondlife (Read error: 110 (Connection timed out))
00:21:00 Join japc [0] (
00:21:06barrywardellmaybe it's cpu frequency related then
00:21:23 Quit japc (Read error: 104 (Connection reset by peer))
00:21:35barrywardellif you comment out that line, what happens?
00:24:05przemhbbarrwardell: I have just comment out the line and are making...
00:25:06 Join bertrik [0] (
00:25:43przemhbbarrywardell: it get us back to the point where it stucks inside root_menu()
00:26:12barrywardellhmm. I'm out of ideas for now until I can test it on my own H10.
00:27:03barrywardellwhich will be tomorrow
00:28:23barrywardellgot to go now. I'll let you know once I get a chance to test it myself.
00:28:52 Quit barrywardell ()
00:31:52przemhbbye all
00:32:00 Part przemhb
00:32:39 Quit mirak (Remote closed the connection)
00:33:45 Join mirak [0] (
00:34:22 Quit iamben (Read error: 104 (Connection reset by peer))
00:35:13 Quit advcomp2019 (Nick collision from services.)
00:35:27 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
00:36:42 Quit Robin0800 (" I love my HydraIRC -> <-")
00:37:40 Join colin__ [0] (
00:39:38amiconnHmm, bit-banging i2c on PP....
00:39:50 Join iamben [0] (
00:40:54bertrikamiconn: mmm? are there targets that do that?
00:41:32amiconnH10 tuner communication, as discussed by barrywardell and przemhb
00:42:14amiconnE, but the port manipulation can be done atomically
00:42:30 Quit Seed (Nick collision from services.)
00:42:34 Join Seedy [0] (
00:42:41amiconnprzemhb shouldn't have asked about atomic and/or macros on arm, but about atomic gpio manipulation...
00:43:43 Quit ender` (" A computer program will always do what you tell it to, and seldom what you want it to.")
00:44:09bertrikI'm working with a LPC2148 ARM7 and it has GPIO set and clear register, very convenient
00:48:00 Join Seed [0] (
00:48:36 Join scubacoles [0] (
00:50:02BigBambiAny manual people here?
00:50:18*n1s is a manual person
00:50:20BigBambiI can build a (clean) pdf fine but the html errors out
00:50:43BigBambiIf I pop it on pastebin could you have a quick look for me?
00:51:26n1sBigBambi: does the html build without the change?
00:51:34*amiconn never got the html manual to build
00:51:42*n1s neither...
00:51:49BigBambiI made changes, then tried compiling - pdf was fine, html failed
00:51:59BigBambiSo I tried a clean check out and html still failed
00:52:16BigBambiSo it isn't just me :)
00:52:36 Quit Shoot ()
00:52:45n1sBigBambi: if the pdf builds fine you should be fine
00:52:53BigBambinls; OK, cool
00:53:05BigBambiI have just expanded slightly the font installation section
00:53:06 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
00:53:14BigBambiAt the moment it just says "install it"
00:53:27BigBambiAnd we have had a couple of questions saying "How?"
00:53:42saratogaopps, accidentally zeros DEV_EN, which does about what you might expect
00:54:34n1sBigBambi: sounds good
00:54:54 Quit OlivierBorowski (Remote closed the connection)
00:55:20BigBambin1s: I've just removed the note about fonts from the installing firmware section and stuck a new section inbetween installing firmware and installing bootloader
00:55:23 Quit davina (Remote closed the connection)
00:55:26amiconnsaratoga: Yes, as it also disables the whole system
00:55:31BigBambiFlyspray coming up
00:55:45 Join TomatoMonster_79 [0] (n=Thunder2@
00:55:46 Part TomatoMonster_79 ("Wrarrrrr!!!")
00:55:46amiconnEven sleep mode on PP5002 leaves 5 clocks running (iirc)
00:56:50amiconnOf course at a very low frequency: 32768Hz
00:58:23 Quit z35 ("Leaving")
00:58:31 Quit scorche|w ("CGI:IRC")
00:59:25BigBambin1s: There you go:
00:59:54n1sBigBambi: I'll look at it tomorrow, thanks
01:00:27BigBambiI hope I haven't cocked up the latex too badly!
01:01:30 Join barrywardell [0] (
01:03:18saratogaamiconn: yeah i figured, i'm just a littl rusty with bitwise operations
01:04:56***Saving seen data "./dancer.seen"
01:04:59 Quit colin__ ()
01:05:13 Join thegeek [0] (
01:05:26saratogacool disabling USB does exactly like i'd expect
01:05:30saratogaamiconn: any specific bits you'd like to see first?
01:05:39saratogai'm going to try them all
01:05:48 Quit Seedy (Read error: 110 (Connection timed out))
01:05:52 Join colin__ [0] (
01:06:07*amiconn suspects there is more to it than just DEV_EN(1)..DEV_EN3
01:06:11 Quit Seed (Nick collision from services.)
01:06:14 Join Seedy [0] (
01:07:08saratogawell if you have another register, let me know and i'll test it after the DEV_EN
01:07:17amiconnThe G5 ROM fiddles with 0x70000028 and 0x70000060 - these addresses are looking suspiciously like inits/resets/enables
01:07:43saratogado you know which bits?
01:07:54*amiconn wonders how many different functional units the PP502x might have
01:08:31saratogaits hard to imagine that the 96 (?) provided by DEV_EN 1-3 aren't enough
01:08:51saratogaunless thats mapped onto the broadcom part or something
01:09:33saratogaalright i'm moving my system into the lab to start taking measurements, i'll check back in a while
01:09:40 Quit saratoga ("CGI:IRC")
01:10:54 Join T0M4T0_Monster_9 [0] (n=blastekk@
01:10:59 Quit T0M4T0_Monster_9 (Excess Flood)
01:15:15preglowargghghg, my linux has gone insane :/
01:17:30preglow800x600, how i have missed thee
01:17:36 Quit bertrik ("bye")
01:18:01amiconnsaratoga (for the logs): On a closer look, those addresses look more like resets and more complex inits to me. They seem to be related to ata somehow - they appear in conjunction with 0xc5000xxx
01:18:24amiconnsaratoga: yeah, but otoh this beast obviously has 128 GPIO ports...
01:20:37amiconnThe broadcom is hooked up to the address & data bus as normal. It has only 3 address bits (8 addresses), mapped to PP address bits 16..18, and a 16 bit data bus
01:21:29 Join saratoga [0] (i=98039b2b@gateway/web/cgi-irc/
01:21:45saratogaamiconn: the sansa battery has 3 terminals, two read negative but are not shorted and the other is positive
01:21:52saratogaany idea what the second negative is for?
01:22:08amiconnSome sense signal most probably
01:22:24saratogaany guess which i should ground?
01:22:34amiconnUnfortunately not
01:22:54saratogai assume grounding the wrong one is harmless enough
01:23:26amiconnDepends on how the protection circuit works. Is the sansa battery user replaceable?
01:23:50 Join Isolinear [0] (
01:24:05 Join darkapostrophe [0] (
01:26:18 Quit barrywardell ()
01:27:24amiconnThe small H10's battery also has 3 terminals
01:28:00 Join iamben_ [0] (
01:28:36 Join T0MA4AT0_M0nst4h [0] (n=HARSKISU@
01:28:40 Quit T0MA4AT0_M0nst4h (Excess Flood)
01:29:53amiconnHmm, but that one has the 2 negative terminals shorted
01:30:08saratogai think teh two need to be shorted to boot
01:30:20saratogaat least it powers off right away with the mid terminal floating
01:32:25 Part toffe82
01:34:39Soapamiconn: exactly as you predicted - 20% improvement in 5th gen runtime with HDD poweroff.
01:35:02preglowsaratoga: found anything fun?
01:35:19amiconnWell, it's an improvement....
01:35:28 Quit scubacoles ("Leaving")
01:35:41saratogapreglow: I can't get it to boot without the battery in place
01:35:47*amiconn wonders what's up with rockbox' main menu
01:35:51saratogathat third pin seems to do something important
01:36:28amiconnIt draws more power than submenus, and power consumption fluctuates
01:36:32preglowthird pin? probably some kind of sense
01:36:34amiconn(measured on 1st gen)
01:36:57preglowwhich explains why it won't boot, it probably thinks the battery is drained
01:36:59saratogai think its a resistive sensor
01:37:09saratoga115k ohms but zero open circuit volts
01:37:11amiconnAh no. Seems to be related to wheel position
01:37:22saratogawill get a 100k resistor and try again
01:37:44Soapthough it also looks like there have been some measurable improvements prior to 15444 as well. My battery is obviously aging rapidly, yet the pre-HDD runtime was also suprisingly long. I'll start the Apple Firmware runtime tomorrow to get a RB%.
01:37:50n1sthat's a pretty big improvement, when I tested runtime with/without hd poweroff on h320 it went from ~13 to ~14 hours ~7-8% improvement
01:38:37preglowamiconn: i wonder if i should even bother trying to test hd power off on nano :>
01:38:39amiconnYeah, but that was still with the extra power sucker that was fixed afterwards
01:38:59amiconnpreglow: Looking through the rom it has no hd power off
01:39:03SoapI haven't thought the whole thing through yet, but does the fact the 60GB 5th gen has 64MB of RAM increase the value of HDD poweroff? Or does it all balance out?
01:39:09preglowamiconn: then i shall most certainly not bother
01:39:12n1samiconn: I'm fairly certain the sucker was fixed before that
01:39:16preglowamiconn: surprising, though
01:39:19amiconnThe nano seems to be the only ipod with a 512KB rom
01:39:36amiconnAll others have 1MB, even the old ones (1st and 2nd gen)
01:39:43 Join Ghwomb [0] (
01:40:00amiconnn1s: I doubt it. 14 hours seems a little low...
01:40:06preglowjmspeex: awake?
01:40:31n1samiconn: it was with q6 vorbis though IIRC so that might be it
01:40:54 Join Shagnar [0] (
01:40:58preglowamiconn: weird
01:40:59 Quit ompaul (Client Quit)
01:40:59 Quit iamben (Connection timed out)
01:40:59amiconnMy old tests were (that was definitely after fixing the otg power sucker):
01:41:10n1sbut anyway impressive improvement for the ipods :D
01:41:29amiconnH340, with hd poweroff: 20:26, without hd poweroff: 15:24
01:41:31nanokSoap: if you mean balance out with the increased disk size, normally it shouldn't: it depends more on other factors. however, in the atypical situation where you seek like crazzy all the time, i guess it would come close to balancing out (at least that's my view at first glance)
01:41:37 Quit Dark_Apostrophe (Connection timed out)
01:41:40 Quit HellDragon (Client Quit)
01:41:45amiconnThat's my usual test set, mp3 lame −−preset standard
01:41:54amiconnAround 200kbps
01:42:32Soapnanok: no I mean - since the disk spins half as often for for ~twice as long, and HDD poweroff timeout is short - does it make any difference on how much juice the HDD consumes.
01:42:51Shagnarhey guys... in 2005 the HDD of my h140 broke... after i just installed the rockbox-early-firmware.. just at this moment i got it running again... rockbox was able to play the first 8megs of the \sample.wav file and it was grandiose...
01:43:02amiconnpreglow: Btw, hd poweroff doesn't help on mini either. Seems those microdrives do auto-sleep like cf
01:43:17nanokSoap: aaahm :). no, it definetly does not balance out like that
01:43:44 Nick darkapostrophe is now known as Dark_Apostrophe (
01:43:51 Quit Ghwomb (Client Quit)
01:43:58nanokSoap: from that point of view, more memory is better. however, when the listening pattern is random (unpredictable), too much memory and buffering will only increase power consumption
01:44:03Soapnanok: I would also think if you seek and skip "like crazy" the double buffer would actually hurt you.
01:44:09 Quit Arathis ("Bye, bye")
01:44:56nanokSoap: so there is a limit, depending mainly on the average size of the media you listen too, in that case, from up to which increasing buffer memory will make things worse, i guess
01:45:06nanokSoap: exactly
01:45:25preglowwrong window
01:45:25amiconnIf you never skip, more memory is better
01:45:33Soapbut that isn't what I was trying to get at. My query was - using the standard battery bench procedures of no skipping - will the 64MB buffer models see more beneft from HDD poweroff than 32MB models, less, or the same.
01:45:42n1samiconn: my tests were run on december 8 last year, several months after the battery sucker was fixed, but I guess it's the decoder efficiency difference
01:45:48amiconnThe benefit should be about the same
01:45:50nanokSoap: however, it shouldn;t be hard to manage this smartly, and buffer in an...adaptive way, so to speak
01:46:31nanokSoap: more beneffit, definetly, if the buffer is used correctly
01:46:50 Part pixelma
01:46:56Soapnevermind nanok - that isn't the question I was asking. :)
01:47:00nanokSoap: spinning the disk once to read more is cheaper than spinning the disk twice to read half as much every time
01:47:04amiconnFilling the 64MB buffer takes twice as long as filling a 32MB buffer. YOu just save one spinup
01:47:27nanokSoap: for that simple question, the answer is at least 99% sure: more memory is better
01:47:35amiconnThat's 2..3 seconds the hd can sleep longer in one "64MB buffering period"
01:48:01Soapamiconn: thank you - that is what I assumed, but wasn't sure if I was missing part of the puzzle.
01:48:11Soapnanok: that still isn't what my query was about.
01:48:15nanokexactly, and the greatest power consumption should be spinning up, not keeping the speed
01:49:11n1snanok: entirely depends on how long the disk is spinning
01:49:16nanokSoap: maybe i am not expressing myself so clear, it is long past my bedtime ;), but i basically said what amiconn explained, but not in such well chosen words as he did ;)
01:49:44*amiconn should go to bed too
01:50:48saratogaugh damn sandisk for putting these pins so close
01:51:23 Join HellDragon [0] (i=jd@unaffiliated/helldragon)
01:52:07amiconnMy battery bench list still has *the* rockbox battery runtime record :)
01:52:10 Join psycho_maniac [0] (
01:53:24 Join T0MA4AT0_M0nst4h [0] (n=aTb-off|@
01:53:26 Quit T0MA4AT0_M0nst4h (Excess Flood)
01:54:08 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
01:54:25 Join Calcipher [0] (
01:55:29Soapand you even have the iPod record as well, I coveted that one.
01:56:11amiconnMy 2nd gen test....
01:56:26Calcipherhey people
01:56:56saratogai can get it to power on the screen for a second or two, but it always powers off
01:57:24Calcipherdid anyone see my comments yesterday, about a problem I encountered, while generating voice clips in cygwin
01:57:53 Quit mirak ("Ex-Chat")
01:58:00saratogacurrent draw seems reasonible
01:58:00saratogai've tried 65kohms, 0 ohms, 100kohms, and 150kohms shorting the sense pin
01:58:02 Join thegeek_ [0] (
01:58:12saratogai'll try 10k
01:58:58preglowwhat's your setup anyway? why remove the battery?
02:00:22 Quit psycho_maniac (" Want to be different? Try HydraIRC -> <-")
02:01:32 Quit colin__ ("")
02:01:50amiconnOh. Didn't even remember that I did such a measurement, but rombox on archos (measured on player) extends battery runtime by 4..5%
02:02:05amiconnThat's even though code runs slower from rom than from ram
02:02:38saratogapreglow: the battery connects to the sansa via pressure from the casing
02:02:38saratogait has no wires, just metal contacts on the side
02:02:38saratogatheres no obvious way to insert an anmeter without cutting traces on the sansa
02:02:38DBUGEnqueued KICK saratoga
02:02:38saratogaso i'm using a lab support with an anmeter in series
02:02:38saratogalab supply
02:03:01Calcipherthe error I recieve after the voice build steps, when I run make voice, is as follows
02:03:08saratogai wonder if the load across the battery changes the impedance seen on the sense pin
02:03:35amiconnCould also be some more sophisticated battery identify thing
02:03:47CalcipherGenerating voice clipserror -2147024894
02:04:46Calcipherfollowed on the next line by "make: *** [voice] Error 141
02:05:20Calcipherthis was selecting english, then selecting sapi, and default lame and sapi settings
02:05:56Calcipherfor the e200R
02:07:44IsolinearIs it just me or have the more recent builds seemed to slow down the iPod 5G?
02:07:59Calcipherbefore that, I was having an issue that Mouser_X was familiar with, and helped me through, that was that ther arm-elf-gcc path was incorrect
02:08:09saratogathis is so weird, the resistance doesn't change under load, so i don't see why the sansa isn't happy
02:08:11 Join sarixe [0] (
02:08:19saratogasame voltage, current, imedance on all pins
02:09:35preglowafaik, there are fancy things like that around that even communicate digitally
02:10:41 Quit Toxicity999 (Remote closed the connection)
02:11:14 Join ghost [0] (
02:11:20saratogaif theres an IC on the battery i'm screwed
02:11:22 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
02:11:33 Quit ghost (Client Quit)
02:11:58preglowi'm quite certain that exists
02:15:45 Quit thegeek (Read error: 113 (No route to host))
02:16:38saratogaGOT IT
02:18:00Mouser_Xsaratoga: What is it you got? I joined a tad late.
02:18:15preglowsaratoga: doing what?
02:19:58 Join webguest93 [0] (i=4c10ea95@gateway/web/cgi-irc/
02:20:00saratogai think the current spikes high enough on the DMM to engage the overcurrent protection
02:20:19saratogait works fine if i use the course current meansurement terminal, but then i don't get accurate enough readings
02:20:32webguest93hi do you have a voice file in spanish for the ipod video 30g 5generation?
02:24:10 Quit spiorf_ (Remote closed the connection) you guys have the voice file?
02:25:44saratogapreglow: i can't work the controls while the sansa is wired up to the supply, so wheres the best place to hard code PP register values into? system-PP5020?
02:26:54 Quit Langly ("I dont have anything to say, thats why I'm quitting in the first place.")
02:27:06preglowsaratoga: what, so it just does fancy writing on boot?
02:27:45webguest93you guys to busy? i can try another time...
02:27:57preglowwebguest93: if someone had it, they would have answered
02:28:09preglowyou can still try another time, though
02:28:19 Quit Thundercloud (Remote closed the connection)
02:29:25 Join homielowe_ [0] (
02:29:34jmspeexpreglow: yes
02:29:44 Quit homielowe (Read error: 104 (Connection reset by peer))
02:29:52 Nick homielowe_ is now known as homielowe (
02:30:36saratogapreglow: I just need a place to stick the DEV_EN = DEV_EN & 0xFFBFFFFF; lines somewhere so i can see what different settings do to power
02:31:27 Quit webguest93 ("CGI:IRC (EOF)")
02:31:30 Join aliask [0] (
02:33:03preglowjmspeex: do you see anything wrong with the fp math in this patch? it applies again rockbox apps/codecs/libspeex dir,
02:33:15 Quit dandin1 ()
02:33:42preglowjmspeex: i've done a comparison too, and that code differs from the old floating point based code by -45 dB, with ony one big peak keeping it from -55 dB
02:33:57preglowsaratoga: well, some place in apps/main.c is probably good
02:34:12preglowsaratoga: just check out app_main() or something, that's what gets called first
02:34:41saratogathat PP file seems to have worked
02:34:57preglowsure, as long as it gets called during startup, which it probably does
02:35:01saratogaoh maybe not
02:35:40saratogai'll try main.c
02:35:43 Join darkapostrophe [0] (
02:39:40saratogapreglow: thanks working good now
02:39:42Calcipherwell if anyone has experience generating voice files in cygwin using SAPI, please lend me a helping hand here, I'll be back later, hittin the gym, with my rb'ed Sansa, enjoyin the hell out of it
02:39:53saratogaUSB uses 1.5ma
02:39:53preglowsaratoga: great, looking forward to seeing results :)
02:40:22CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
02:40:22*Calcipher checks manual for new keymaps
02:40:58saratogaonto the next bit
02:43:32 Join ddalton [0] (
02:44:24ddaltondo I just need to run the rockbox sh script on linux to get all the compilers and stuff?
02:46:21 Nick fxb is now known as fxb__ (
02:47:49preglowjmspeex: i could probably have used more precision for the iir filter state and stuff, but 16 bit seems to be ok
02:49:26saratogathe sansa will happily draw power from the USB cable if I disable USB detection and remove the battery
02:49:37saratogadoes USB have a +5 line thats seperate from data?
02:49:51saratogathat would be MUCH easier to tap into then the battery pins
02:50:14n1ssaratoga: usb is two power lines and two data lines
02:50:34saratogathen i should be splicing a USB cable
02:50:54preglowsaratoga: it does
02:50:58preglowwhat he said
02:51:00saratogahmm bit 31 seems to do nothing that i can tell
02:51:12saratogarockbox works normally and the current draw doesn't change
02:51:15jmspeexpreglow: is the -45 dB due to the 0.01 on silence or due to the fixed-point?
02:51:36preglowjmspeex: it might be due to 0.01, floating point roundoff due to e_tot, or fixed point
02:51:50jmspeexpreglow: never, ever use the >> or << operators on spx_word* types
02:51:50preglowi dunno, really
02:52:10preglowjmspeex: do i do that?
02:52:33jmspeexyes, at the end. It's bad even on a fixed-point path
02:52:43jmspeexalso, don't use "int" for sign.
02:52:56preglowforgot about that at the end
02:52:57 Quit Dark_Apostrophe (Connection timed out)
02:52:57 Nick darkapostrophe is now known as Dark_Apostrophe (
02:52:58preglowwhy not use sign?
02:53:08preglowint for sign, i mean
02:53:56preglowi don't do << or >> on spx_word types, i do it on an int
02:54:11preglowin a code path which is always fixed point
02:54:20jmspeexint is platform-specific.
02:54:26preglowsure, but it's always signed
02:54:29jmspeexYou need to use spx_word16_t or spx_word32_t
02:54:52preglowit can either be 1 or -1, which any int should be able to accomodate, but sure, it's not a big deal anyway
02:55:01jmspeexspx_word16_t is either an int16 or a float depending on whether you enable fixed-point or not.
02:55:16jmspeexAlso, you shouldn't be using the * and +, but rather the ADD and MUL operators.
02:55:38preglowhmm, where?
02:55:54jmspeexone immediate benefit of doing things properly is that compiling with fixed-point-debug will warn if you get overflows
02:56:18preglowi think i've always used macros in my code, the floating point path remains for reference and is your code, i'll remove it when commiting
02:57:19saratogabit 30 uses 300uA but does not disable anything obvious that i can see
02:57:34preglowok, so i should make tmp a spx_word16_t too?
02:57:47preglowsaratoga: could you stuff this in a doc somewhere? wiki?
02:59:04preglowjmspeex: i don't really see a point in bothering with the macros when i'm in a non-critical fixed point only code path, though
02:59:08jmspeexpreglow: you should probably assign tmp to a word16 variable just to make it obvious what you're doing
02:59:15preglowyeah, done
02:59:56preglowmade it exp_exp(MULT16_16(sign, SHL16(tmp, 9)));
02:59:58jmspeexpreglow: The idea of the macros is that they can be adapted to any platforms. Also, they allow testing for overflows
03:00:14preglowoverflow testing is good
03:01:01jmspeexpreglow: looks goot
03:01:52preglowi'm wondering if the innerloop smooth_left/right assignments can overflow
03:02:20preglownah, they should _just_ fit within the limits
03:02:34jmspeexpreglow: if you look at fixed_debug.h to have an idea of the tests I do
03:02:43preglowdo you think the rounded shifts are necessary?
03:02:49jmspeexcareful about 32768 too
03:03:43 Quit GodEater_ (Read error: 110 (Connection timed out))
03:03:51preglowwell, neither of them should be negativ
03:04:07preglowsmooth_left or e_left, so -32768 shouldn't be a problem
03:04:25preglowhmm, the scaling constants sum to 32768, probably
03:04:59preglowperhaps i should just go with the commented out line, that will never overflow, but is a slightly badder approximation to 0.98 and 0.02
03:05:01***Saving seen data "./dancer.seen"
03:05:17preglowi should just test...
03:05:18jmspeexpreglow: BTW, there's a useful macro named Q15_ONE that's often useful
03:05:32jmspeexis defined as either 1.0f or 32767
03:05:50*Nico_P has hackish AA!
03:06:17preglowjmspeex: oh, didn't know about that one
03:06:44preglowhmm, and can't use it here, i need Q16_ONE
03:07:05 Quit Calcipher (Read error: 104 (Connection reset by peer))
03:07:27preglownah, it'll never overflow
03:07:59jmspeexQ16_ONE??? doesn't make sense for signed
03:08:26preglowjmspeex: for 32 bit Q16 ints it does, heh
03:08:32jmspeexpreglow: also, you need to be careful with rounding shifts because of the way they're implemented.
03:08:54preglowyeah, know, but the rounding const is too small to matter here
03:09:12jmspeexFor example, PSHR16(32767,2) will actually cause an overflow because it adds 2 to 32767 before shifting
03:12:32preglowi have no idea if rounding is really necessary here, i just added it to be sure
03:13:03rasherNico_P: I'm almost embarrassed to admit that I'm quite excited about this
03:13:31Nico_Pit's hackish on the WPS side, but on the playback side it's almost complete
03:13:54*aliask likes AA
03:14:21*Nico_P is really looking forward to being able to finally close FS #3045
03:14:32preglowjpg loader in the core! :D
03:14:57ddaltonwhat tools do I need on my system for
03:15:10ddaltonit won't work for me
03:15:14jmspeexpreglow: I think it's probably fine without the rounding, though it doesn't hurt with it.
03:15:54rasherpreglow: couldn't it be loaded as a plugin, on demand?
03:16:01 Join kingwen [0] (
03:16:11rasherNot a regular plugin of course, but plugin-ish
03:16:17jmspeexpreglow: there's a problem with e_ratio_quant. You're not allowed to use QCONST16() in a definition like that because it uses floor()
03:16:24n1sddalton: a bunch but starting out with a working native compiler is good :)
03:16:46rasherddalton: on a Debian system, apt-get install build-essential
03:16:50Nico_Prasher: maybe, yeah... though I think jpeg in the core wasn't as badly received as I thought last time it was discussed
03:16:50ddaltonn1s: my system has gcc, mutt, links, emacs and apt and thats about it.
03:16:53preglowrasher: let's not call plugins all the time
03:16:58ddaltonwhat should I get?
03:17:00preglowrasher: it's a horrible hack
03:17:04ddaltonoh and festival and brltty
03:17:08jmspeexpreglow: sorry, it no longer uses floor(), but I'm not yet 100% sure it's OK. Any idea?
03:17:25n1sddalton: as rasher said build-essential package is important, what error do you get?
03:17:26preglowit should be very ok
03:17:46preglowjmspeex: at least i can see nothing wrong with it, it's just reduced to a ton of constant folding which should be resolved real time
03:17:53preglowjmspeex: we have several other codecs which does this
03:18:24preglowjmspeex: as a matter of fact, i was about to ask you why you don't use this technique in other parts of speex :) it's easier than duplicating tables
03:18:29rasherpreglow: well, if it's the only way to get jpeg loading accepted, it's better than nothing. But I don't see why jpeg in the core is so unwanted - is binary size an issue on any relevant target?
03:18:35kingwen :o
03:18:37preglowrasher: hell no
03:18:42preglowrasher: jpeg in the core is the way to go
03:18:42ddaltonn1s: ill just check. Do you know how I get the build stuff on debian?
03:18:45jmspeexpreglow: so you're sure it's perfectly C89-compatible?
03:19:12preglowjmspeex: i think it is c89 compatible, but it might be highly ineffective with compilers that don't optimize away the float code compile time
03:19:15n1sddalton: <rasher> ddalton: on a Debian system, apt-get install build-essential
03:19:16Nico_Plet's stick to bitmap for now gentlemen ;)
03:19:21preglowjmspeex: and i am by no means certain it _is_ c89
03:19:30Nico_PI remind you that we don't even have AA yet
03:19:38ddaltonI need bzip apparently whats that by the way?
03:19:57aliaskIt's a compression format, similar to zip
03:20:04Nico_Panyway, bed time now
03:20:06n1sddalton: google is your friend
03:20:15*n1s goes ZzZz too
03:20:17 Quit n1s ()
03:20:21 Quit Nico_P (Remote closed the connection)
03:20:30jmspeexpreglow: well, if it's in a const struct definition, there's no way the compiler can compute it at run-time, so either it accepts if or it reports an error.
03:25:37preglowit can if it's not a static const struct definition, afaik
03:26:09preglowthe same way const arrays in functions can end up being constructed for each function call if they're not declared static
03:26:29 Join Gigante [0] (i=Gigante@
03:26:57Gigantei need help with a certain subject
03:27:23aliaskGigante: What is that subject?
03:27:40ddaltondoes anyone know what I press for all the compilers so I can build any target?
03:27:53ddaltonI just am having a bit of trouble with brltty.
03:28:02Gigantei work a part time job at an airport
03:28:12Gigantefound me a 30 gb ipod
03:28:30Gigantebought a cvable but itunes doesnt recognizes it
03:28:51Gigantethe battery is full
03:29:15Gigantei trieds to format it on win xp
03:29:17preglowjmspeex: anyway, i need to go to bed now, talk to you later
03:29:32Gigantebut when i click f drive it freezes
03:29:48aliaskPersonally I think you should return it to lost property at the airport, but it sounds like the hard drive isn't formatted correctly.
03:29:56jmspeexpreglow: OK. But overall, your patch looks fine. It was just formatting details to make it like the rest of the code. Thanks for that
03:30:13Gigantei did
03:30:24preglowjmspeex: i'll fix up the rest of stereo.c and give you a proper patch against speex svn soon
03:30:33Gigantethey wait a week and if not claimed it's yours
03:30:41jmspeexpreglow: thanks, g'night.
03:30:50aliaskWell then, nice score
03:31:13Giganteit's a little bit scratched but i dont complain
03:31:32Gigantecan i reformat the ipod and load rockbox?
03:31:44aliaskiTunes should be able to restore the iPod
03:31:59Giganteitunes doesnt recogmnizes it
03:32:13aliaskI'm not sure if it's safe to format an iPod though, because the firmware is stored on the harddrive
03:32:19 Quit Shagnar ("( :: NoNameScript 4.1 :: )")
03:32:39Gigantehow to change firmware then?
03:33:11ddaltonaliask: couldn't you use ipodpatcher to restore it later?
03:33:17ddaltonif your on win of course
03:33:49Giganterigth now is in disk mode
03:34:17aliaskGigante: Can the iPod load normally?
03:34:30Gigantestops when apple shows
03:35:12aliaskSounds like the firmware on the disk is borked. Disk mode is in flash though. You should take a look at restoring the iPod first, and then installing rockbox
03:35:45Gigantehow i restore it?
03:36:51ddaltondoes voice building require anything else apart from the targets compilers?
03:37:31 Quit homielowe (Read error: 104 (Connection reset by peer))
03:37:43 Join homielowe_ [0] (
03:37:48 Nick homielowe_ is now known as homielowe (
03:45:22 Join webguest95 [0] (i=dce9bdd9@gateway/web/cgi-irc/
03:46:34 Quit webguest95 (Client Quit)
03:46:59ddaltonhow long should the compilers take to install?
03:47:15ddalton800 mhz processor 256 mb ram. 2 hours?
03:49:38 Quit kingwen (Read error: 110 (Connection timed out))
03:53:28Gigante0i tried to restore and received this message: ipod could not be restored an unknown error ocurred 1418
03:54:17 Quit ddalton (Read error: 104 (Connection reset by peer))
03:54:27aliaskHrm, odd. Might need to contact apple for support
03:56:13 Join psycho_maniac [0] (
03:57:28 Join ddalton [0] (
04:05:17 Quit DerPapst ("So Long And Thanks For All The Fish!")
04:12:43 Join ruinedtwice [0] (
04:20:12aliaskZagor: (for the logs) Those USB registers seem to have some kind of voodoo magic protection on them, as soon as I try to read them it locks up. Other registers read fine...
04:25:01Mouser_Xlostlogic: I have a question regarding buffering. It regards ADX, but what I'm thinking is usable for MP3s as well, if I'm understanding correctly (actually, it's probably the other way around - MP3s would use it, and it might be useable for ADXs).
04:25:53lostlogicI'll try not to ignore you.
04:26:06Mouser_XQuestion: If someone has a playlist of 9 songs (large enough to cause rebuffering), 3 of which are the same file, that file will still be rebuffered, correct?
04:26:21lostlogicyes, Nico is working on enabling smarter buffering though.
04:26:36lostlogicthe full implications thereof are complex when it comes to buffering shrinking and suchlike
04:26:56Mouser_XCould that be applied to ADXs as well? To keep them in the buffer, so that they loop properly?
04:27:41lostlogicyeah, if you had a playlist of the same ADX twice and Nico does what he's talkin' about it would enable that ADX to not partially unbuffer and thereby not feck itself up
04:27:43Mouser_XObviously, I'm not the one to say, but if the buffer stuff is "smart enough" to keep the MP3 in the buffer, wouldn't that make it easier to do the same for ADX?
04:28:07lostlogicbut we do seem to need a solution for files that are guaranteed not to be >= bufsize and which we need to keep all of on buffer for the duration if its playing, instead of shrinkign it
04:28:11lostlogicat least that's my current take
04:29:23Mouser_X^ That would fix ADXs, if I understand correctly, right? (I'm not completely sure what ">=" means. It's been to long since I did math...)
04:30:04lostlogicgreater than or equal to...
04:30:06lostlogicand yes.
04:34:08ddaltonenouf: ok so how do I use that?
04:34:25ddaltonoops wrong channel sorry
04:54:16 Quit Gigante ()
04:57:57 Join Llorean [0] (n=llorean@
04:59:14 Join TMM [0] (
05:05:04***Saving seen data "./dancer.seen"
05:07:46 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
05:08:02 Join miepchen^schlaf [0] (
05:15:14 Quit bb (Nick collision from services.)
05:15:21 Join bb_ [0] (n=bb@unaffiliated/bb)
05:24:38 Quit homielowe (Read error: 104 (Connection reset by peer))
05:24:58saratogadid amiconn ever say what the address for dev_en3 was?
05:25:11saratogathe one in the wiki is the same as the CACHE_CONTROL define
05:25:21saratogaso i'm assuming thats a mistake unless they're the same thing
05:26:22 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
05:28:29 Quit psycho_maniac (" HydraIRC -> <- It'll be on slashdot one day...")
05:29:56 Join homielowe_ [0] (
05:30:01 Nick homielowe_ is now known as homielowe (
05:42:27 Quit ruinedtwice ()
05:52:35Mouser_XI've never done a battery bench on my Gigabeat, but so far, it's been running for almost 15 hours, and still has 26% left (according to Rockbox).
05:52:51 Quit SirFunk (Remote closed the connection)
05:55:46LloreanThe gigabeats seem to land in the "decent+" battery life category. :)
05:56:28 Join RoC_MasterMind [0] (
05:57:37Mouser_XYes, I've noticed it's 15+ hours. I've just never bothered refining it more than that.
05:58:42Mouser_XIf I can get 17 hours out of it, then that should mean I can have it play for 2 days, without recharging, at work (8 hours).
05:59:03Mouser_X(17, because I tend to watch videos during breaks.)
05:59:51 Quit sarixe ("(EE) Failed to load "quit" module")
06:01:02saratogaturning off most of these registers doesn't seem to do anything to rockbox
06:01:18saratogathough i have to wonder if i'm missing some ill effect somewhere
06:06:03LloreanMouser_X: Video takes up a LOT of juice. I think I timed something like 5 or 6 hours at 16x9 for video
06:06:18LloreanThough that was when it wasn't as optimized as it is now.
06:06:43PaulPositionWhat's the runtime for the original GB firmware? Just setting myself up for unattainable hopes regarding the pp targets.. :p
06:07:21LloreanPaulPosition: Rockbox only really matches the PP targets, or maybe gets 5-10% more. There aren't any reliable direct comparisons.
06:07:48LloreanErr, only really matches on the Gigabeat targets
06:08:03LloreanOn the PP5002 targets we do already get some more too, there should be benches somewhere
06:09:04PaulPositionLlorean - I know.. Some targets match okay, some are far from there (like small H10), etc.. I just wanted to day-dream.. :)
06:09:33saratogaLlorean: I found a few PP registers that gate small amounts of power so far
06:09:38saratoganot sure what they do
06:09:55saratogado you think i could get away with committing a few after some testing and just waiting to see if anyone complains?
06:10:07Mouser_XLlorean: Yes, I'm aware that video does nasty stuff to battery life. However, today I watched at least 1 full hour of video (today I had two 15 min. breaks, and an hour lunch. I watched some TV shows during that time).
06:10:41Lloreansaratoga: If you say "These should improve battery life. If there are any unexpected side effects, please file a proper bug report" I think you'll be fine
06:10:45PaulPosition(note I said hopes and not expectations. I think RB is amazing all righ, as it is now, on my dap)
06:10:59LloreanMouser_X: One hour of video is probably about the same as 2.5-3 hours of audio.
06:11:29Mouser_XI couldn't say. As I said, I never bothered checking before.
06:11:57LloreanCould be better now, but as the main problem is the LCD backlight, it's just kinda a given battery life during video has to suck
06:12:30Mouser_XOn a "normal" day, I get at most 45 min. of video in (usually less). My backlight is at lvl. 7.
06:12:49LloreanMy test was with full backlight, I suppose.
06:12:54 Join SirFunk [0] (
06:12:57LloreanI also haven't tested recently, or with fullscreen video.
06:13:02LloreanWhich takes more CPU to decode, at least
06:13:13Mouser_XVery true.
06:13:46Mouser_XI can't go much darker on the backlight. It's already about as low as I can go, and keep it reasonably visable.
06:13:49LloreanUsually I don't pay attention at all to my battery level, because I know there's almost nothing I can do in the normal use time that will wear it down before I charge.
06:14:04LloreanI really don't understand why people are unwilling to charge once a day though, and want to stretch it through as many days as possible.
06:14:12Mouser_X(For videos, that is. I *could* go lower for normal use, but it starts getting difficult.)
06:14:34 Join webguest70 [0] (i=dead5cb6@gateway/web/cgi-irc/
06:14:50Mouser_XWell, I just don't want to worry about having to replace the battery.
06:15:07Mouser_XAnd no, I have no idea how soon/far away that would be.
06:16:37 Quit webguest70 (Client Quit)
06:16:55PaulPosition(only speaking for myself, for 3 and a half hours of runtime *without* any interaction means more than 'charge once a day'... It's *just* enough for what I use my dap for, but it happened a few times it would die on me in the bus...)
06:18:01LloreanMouser_X: If it's li-ion charging it daily shouldn't be a problem...
06:18:25 Join Maxexcloo [0] (
06:19:00PaulPositionLlorean - You think, with LiIon, it makes sense for me to do, once every two-three weeks, a full discharge/recharge ?
06:19:33LloreanPaulPosition: With lithium based batteries there should be minimal-to-no memory effect. It should be safe to charge it whenever you feel like it.
06:19:47LloreanIn fact, long charges are unhealthy because they tend to heat the battery more, and a hot battery loses life.
06:20:02Mouser_XI did not know that.
06:20:26PaulPositionHmm, interresting.. Though I reckon charging through usb can't be that devastating.. But what do I know. (oh, I know : nothing)
06:20:33Mouser_XThat probably explains why I can't find any of the 15 min. rechargeable NiMH batteries I like.
06:21:03MaxexclooI have a question
06:21:09Mouser_X(They get really hot when charged in 15 min. Sometimes hot enough that I have to juggle them...)
06:21:23MaxexclooHow hard would it be to make gba games play on rockboy?
06:21:34Mouser_XVery, very, very hard.
06:21:52LloreanMouser_X: NiMH batteries are different. They're healthier if you do full charge/discharge cycles, though I believe quick-chargers are rough on them
06:22:06LloreanMaxexcloo: Can you magically make the processor about 5 times the speed it is now?
06:22:40 Join linuxstb_ [0] (
06:23:15MaxexclooI gtg
06:23:21 Quit Maxexcloo (Client Quit)
06:27:20PaulPositionWell, thanks all, good luck, and I'll come tomorrow, download a build and test (or if no commit by Saratoga, download the diffs, patch and compile and then test) for any sort of bug/freeze...
06:28:25PaulPositionand then pester the channel with bug reports.. (dunno why, that doesn't sound like something usefull, what with being a pain in the arse...)
06:29:17 Quit PaulPosition ()
06:34:58 Quit linuxstb (Read error: 113 (No route to host))
06:47:42 Join hannesd_ [0] (
06:51:56 Quit RoC_MasterMind ("Leaving")
06:57:48 Part Mouser_X
07:03:09 Quit hannesd (Read error: 110 (Connection timed out))
07:03:10 Nick hannesd_ is now known as hannesd (
07:05:06***Saving seen data "./dancer.seen"
07:24:22saratogawell, disabling all the DEV_EN1,2,3 bits that aren't needed to playback audio on the sansa drops current consumption from 59.5 to 52.3 ma at idle
07:24:34saratoganot exactly impresive, but its a good start
07:25:10LloreanIsn't the Sansa already much closer to OF performance as it is?
07:25:10saratogathat said, I did notice 1 bit that substantially increased power, so i have to wonder if turning some of the others on could reduce consumption further
07:25:24saratogayeah its pretty close
07:25:28saratogathat will make it closer
07:25:34saratogabut still not up to it I think
07:25:43saratogadamn it i should have metered the OF while I was there
07:26:50LloreanIf we're getting ~15 hours, and assuming approximately consistent ratios, shouldn't we now get about 17? What's our goal?
07:27:17saratogawe get 16 now so i think this will put us to 18 hours
07:27:32saratogasandisk claims 20, though who knows
07:27:38Llorean18 and a quarter even, assuming the math holds.
07:27:51saratogaof course i'm not sure what the load current is
07:28:01saratogatheres no easy way to measure it because of the frequency scaling
07:28:15saratogaso i odn't know what 7 ma is as a percentage of typical load
07:28:57LloreanD'you know what was done for the PP5002 targets?
07:29:08saratogabasically the same thing i did here
07:29:33saratogatheres a good chance theres more registers on the 50xx that no one has found yet
07:29:41LloreanD'you know how much current consumption changed with his changes?
07:29:48saratogai'm not sure
07:29:55saratogaamiconn didn't have a meter hooked up
07:30:32saratogai still wonder if we properly sleep the second CPU
07:30:55saratogathat might be a good bit of it, though with the smp changes, maybe it doesn't matter much
07:31:24LloreanWasn't there a time when we were running all of everything on both cores?
07:31:28LloreanOn accident?
07:31:31 Quit criznach ("ChatZilla [Firefox]")
07:31:36saratogai think so
07:31:39LloreanOr was that just a "theory of what might've been going wrong"
07:31:59LloreanBecause I seem to recall battery life never significantly decreasing during dual core tests, suggesting at least a bit that we never slept it right
07:32:15saratogathat seems likely
07:33:29LloreanThen again the PP5002 improvement suggests we either don't need to, or it's not a significant enough problem to matter.
07:34:48saratogathats true
07:35:04saratoga(for the logs) heres the changes I made to get those 7 ma
07:35:26jhMikeSsaratoga: the cop is woken for ticks or thread wakeups otherwise it's asleep just like the CPU. Nothing was overlooked or shortcuts taken with dualcore in any regard.
07:35:34saratogai'm too tired to write up a proper patch
07:35:39saratogaok thanks
07:36:26saratogaunfortunately, at least 1 ma of that savings is from disabling USB, which is ok on the sansa since we don't properly detect it anyway, but needs to be setup on the other targets so that it polls the port every so often
07:37:21jhMikeSwe need to not poll anything (usb ticks) and use interrupts from the proper source
07:37:23saratogajhMikeS: one question
07:37:41saratogadisabling one of those bits causes a stack overflow error
07:37:50saratogaany idea what hardware that could be?
07:38:41jhMikeSwhich bit?
07:39:17saratogabit 1 in dev_en3
07:39:23jhMikeSI should look at the pastebin?
07:39:34saratogawhich is labeled cache_control in teh code somewhere else
07:39:42saratogaif you want, its not very interesting
07:40:20saratogacache_priority i mean
07:40:43saratogaeh i'm going to bed, if anyone comes up with more registers to test let me know
07:40:52jhMikeSwell, I can't say why. the bits set are as OF did it.
07:41:11 Quit saratoga ("CGI:IRC")
07:55:07 Join mrkiko [0] (
07:56:04mrkikoHi all!
07:56:19mrkikoWhy does rockbox have so much more performances when deactivating voice?
07:56:25mrkikoIt didn't crash, it buffers very very well
07:56:35mrkikoIt simply runs faster!
07:56:41mrkikoI can't get it to crash
08:04:50 Join ender` [0] (
08:18:12 Join criznach [0] (
08:28:23 Join Rob222241 [0] (
08:39:49 Quit aliask ("ChatZilla [Firefox]")
08:46:41 Quit Rob2222 (Read error: 110 (Connection timed out))
08:48:32 Quit EnterUserName (Read error: 104 (Connection reset by peer))
08:52:39 Join daurnimator [0] (i=daurn@unaffiliated/daurnimator)
08:59:24amiconnLlorean: I did (and still do) have an ammeter hooked up to pp5002
09:00:52 Join spiorf [0] (n=spiorf@
09:01:35Lloreanamiconn: I was just curious how Saratoga's changes are relative to your PP5002 changes.
09:02:09LloreanHe said it was a 7ma reduction in consumption.
09:02:19amiconnThe differences on pp5002 were much higher
09:02:49amiconnBut then the Sansa is PP5024, which is already pretty efficient
09:03:09LloreanDid you ever measure which difference was due to HD-related changes, and which weren't?
09:03:45LloreanBut yes, the 5024 seems to be the closest to OF of our inefficient PP targets.
09:04:14amiconnsaratoga's experiment should be repeated on a pp5020 target
09:05:06mrkikoHi Llorean!
09:05:11***Saving seen data "./dancer.seen"
09:05:52amiconnThe biggest problem for such measurements is the ammeter hookup. On 1st/2nd gen it was quite easy compared to later PP targets I have
09:07:05 Join bertrik [0] (
09:07:28LloreanStill, better battery life is better battery life, whether little or lots.
09:07:47amiconnThe old ipods just require to pop off the back cover in order to get to the battery, and the battery uses wires. So I just cut one wire and soldered 2 longer ones into place, where I can hook up my ammeter
09:08:13amiconnBut the mini must be dissected to reach the battery, and then it can't be operated...
09:09:16amiconnAnd the G5 has the battery connected via a flexible pcb strip - cutting that is possible, but how to repair it afterwards?
09:09:36LloreanWhat about the G4?
09:09:47amiconnI don't have one
09:10:34LloreanAh, right.
09:11:55amiconnThe H10 has a user replaceable battery, connected via 3 metal spring contacts
09:12:10LloreanSo that's a better candidate from the sound of it.
09:18:18 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
09:18:27 Join ackbahr [0] (
09:19:03 Part Llorean
09:21:50 Quit ackbahr (Remote closed the connection)
09:36:29 Quit linuxstb_ ("ChatZilla [Firefox]")
09:36:40 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
09:38:14ShaidLlorean: G4 ipod batteries are easily replacable, it's just a simple 2 wire plug connected to the board.
09:38:17Shaidoh, he left.
09:38:38 Join kubiix [0] (
09:38:56amiconnThe important point here being wires, not that it's replaceable
09:39:09amiconnThe G5 battery is also plugged, bu tdoesn't use wires
09:39:38amiconnSo a G4 would be a good target to do power measurements for PP5020...
09:43:02mrkikoI guys! As already said by me some hours ago, I noted that disabling voice rockbox will never crash...
09:43:27mrkikoThe problem of a key release not being detected is triggerable again less easily, butwitout voice it's absolutely not.
09:43:50mrkikoAnd, strangely, without voice I noted much more performance in buffering terms...
09:44:39linuxstbamiconn: I haven't even managed to open my ipod Color... I don't own a soldering iron or any kind of meter either. But I would be happy to lend it to someone.
09:45:29 Join daurn|laptop [0] (n=fake@unaffiliated/daurnimator)
09:45:44linuxstbmrkiko: You're thinking in reverse - Rockbox without voice is "normal", enabling voice causes lots of issues.
09:46:01mrkikoI know :(
09:46:21mrkikoThe problem is that I installed rockbox with voice files from the start of the story.
09:46:40mrkikoI will do without vooice with not many problems...
09:47:20linuxstbAlthough it could be useful if you search the bug tracker for the problems you're experiencing, and file bug reports if they're not already there.
09:47:22*amiconn uses voice on H340, and it's usable
09:47:50mrkikoI don't know of anyone having problems with voice in any other target different than mine - so I expect I should wait someone cleaning up it... I'm trying to understand slowly the code to help eventually...
09:47:50amiconnSure, there are quirks, and I keep moaning about them. Unfortunately I'm not much into the swcodec voice code
09:48:25pixelmaI haven't noticed any major problems with voice on my M5 (which is a vit comparable to the H300) but I admit that I didn't use it very often lately
09:48:37linuxstbMoving the voice codec into the core should hopefully help with reliablity - at the very least it should simplify the code.
09:48:39 Join SkinInd95 [0] (
09:49:11 Quit SkinInd95 (Client Quit)
09:49:13mrkikoCan someone try to trigger fs#8003 scrolling very quickly the main menu?
09:49:16mrkikoespecially amiconn
09:49:27mrkikoso I can detect if it's a problem only mine or general.
09:50:42mrkikoI'm using R15552
09:51:50 Join przemhb [0] (
09:53:37mrkikosomeone should warn me if I'm too stressing :)
09:54:55 Quit kubiix ("Miranda IM! Smaller, Faster, Easier.")
09:58:31 Quit criznach ("ChatZilla [Firefox]")
10:05:20amiconnmrkiko: I can't make it crash by mvoing more or less quickly through the display settings
10:05:59amiconnThis isn't the latest build though. r15312
10:06:04 Quit ATravelingGeek (Read error: 110 (Connection timed out))
10:06:54 Join ATravelingGeek [0] (n=ATG@pdpc/supporter/student/ATravelingGeek)
10:08:10mrkikoI made it crash now - so may it an hardware problem?
10:08:27mrkikoEven ddalton has it
10:08:47amiconnWell, updating all my targets is time consuming... but that build must be replaced as its older than a week. I will retry with a current build later
10:09:00mrkikobut it works only on the main menu, the one saying "files" "database", ..., "settings"
10:09:00 Quit bertrik ("bye")
10:12:18 Join ackbahr [0] (
10:13:30amiconnmrkiko: In what way does it crash? Just freeze, or an error message on the display?
10:13:31 Join GnosticE [0] (
10:17:52mrkikoAmiconn: freeze
10:18:00GnosticEI've got a problem with rbutilqt 1.0.2. When it starts it complains about "Network Error: Proxy Authentication required".
10:18:44*mrkiko loves the new boot-loader
10:18:49GnosticEI don't use a proxy and the No Proxy radio buitton is checked in the config screen. v1.0.1 works fine when I went back to that. Any ideas??
10:20:38GnosticEs/proxy/authenticating proxy/
10:23:45 Quit daurnimator ("Cyas later...")
10:26:17mrkikoamiconn: have you got it crash?
10:28:38 Join merbanan [0] (n=banan@
10:31:25Shaidamiconn: I've got a g4 I don't want if you want to cannibalize it for testing
10:31:42Shaidit has a pretty much dead screen.
10:37:08GnosticEGnosticE has to leave
10:37:24 Quit GnosticE ()
10:43:53 Join Thundercloud [0] (
10:48:04 Join stripwax [0] (
10:49:19 Join barrywardell [0] (
10:51:36 Quit barrywardell (Client Quit)
10:54:16mrkikomrf = ?
11:04:22 Nick parafin|away is now known as parafin (
11:05:12***Saving seen data "./dancer.seen"
11:07:54 Join JdGordon [0] (
11:09:06 Nick fxb__ is now known as fxb (
11:11:32markunmrkiko: just some sounds he makes from time to time :)
11:12:57mrkikomarkun. ok, thank you :)
11:20:58amiconnmrkiko: I can't make the H340 crash in the menu with voice enabled, current build or old build
11:23:05amiconnDoesn't matter whether music is playing in parallel or not, and also not whether I move at a speed that still makes the voice try to speak, or really fast so that voice doesn't speak
11:24:07amiconnWhat I noticed though is that shutup() doesn't work when music is playing
11:24:31amiconnThat probably has to do with the voice mixing
11:24:43 Join davina [0] (
11:24:51 Quit BigBambi (Read error: 113 (No route to host))
11:25:08mrkikoamiconn: strange...
11:25:20mrkikoamiconn is your a USA or non-USA h340?
11:25:30amiconneuropean model
11:25:33 Join BigBambi [0] (n=alex@rockbox/staff/BigBambi)
11:26:00amiconnBut that shouldn't matter at all
11:26:01mrkikoamiconn: mhm...
11:26:17mrkikoamiconn: have you tried in the main menu? In my case it happens only when music is off
11:26:21mrkikook - player crashed again
11:26:43amiconnI tried with music playing and not playing, in main menu, display settings, and sound settings
11:26:53amiconnMoving at various speeds, up & down
11:27:12mrkikoamiconn: so I don't know what to say...
11:27:32amiconnMaybe you're using features I don't use
11:27:42stripwaxmrkiko - you've tried resetting your settings to rockbox defaults?
11:27:57mrkikoNo - infact I'm doing it now
11:28:20mrkikoamiconn: my equalizer was on
11:28:25stripwaxsave your settings first if yuo want them back :)
11:28:27mrkikoamiconn: dithering was off; crossfeed was off;
11:28:59mrkikostripwax: I saved them some time ago probably :) aniway no problems
11:29:16 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
11:29:32mrkikobeep was on but this doesn't matter - it happened in various circumstances even with beep on
11:29:53 Part daurnimator
11:33:51 Join barrywardell [0] (
11:33:52*mrkiko unhappily crashed the player
11:34:05mrkikoOk - my player may be damaged in some ways asthe ddalton one...
11:34:21mrkikoOr may be we have some different revisione of some component
11:34:47 Join n1s [0] (
11:35:39 Join Domonoky [0] (
11:37:47 Quit barrywardell (Client Quit)
11:39:53JdGordonbah.. yesetrdays commit went thourgh without the message? wtf?
11:41:16stripwaxmrkiko - just a thought, checkdisk?
11:41:50mrkikostripwax: I'm a Linux user; fsck.vfat -V -v -r says all is ok
11:43:15 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
11:44:17mrkikostripwax: the fact is that the player never crashed within the OF
11:44:29mrkikostripwax: so I don't expect hardware problems...
11:44:46mrkikostripwax: I tried to respect as closer as possible instructions on the original iRiver manual
11:48:23linuxstbn1s, Nico_P: Have either of you tried the "merge0.cpp" program to inject code into the Gigabeat S nk.bin ?
11:48:25JdGordon:'( no petur
11:48:39Nico_Plinuxstb: no
11:49:47linuxstbDo either of you know where the latest version is? The only one I can find is in MergeInject_v0.2.rar, but the wiki seems to be saying that's out of date.
11:51:45Nico_Plinuxstb: I think you have to compile it
11:51:58linuxstbOf course, but I can't see the source...
11:52:27linuxstb(apart from the version in MergeInject_v0.2.rar)
11:52:35 Join einhirn [0] (
11:52:48 Quit Domonoky (Read error: 104 (Connection reset by peer))
11:52:53Nico_Poh I see what you mean :p
11:53:29linuxstbThe wiki page first says "see below", and then says "see above", with no links to merge0.cpp in between...
11:55:18Nico_Plinuxstb: doesn't version à.2 work?
11:56:25Nico_Por 0.2
11:57:05linuxstbIt doesn't seem to, unless I'm misunderstanding what it should be doing. Running it with the S bootloader (about 32KB) only appends about 92 bytes to the end of the file, and there are no other differences between the original nk.bin and the output file.
11:58:25 Join MouFou [0] (
11:59:02Nico_Plinuxstb: Release.7z has two different versions of merge0.cpp
12:00:11Nico_Phmm actually they don't seem to be what you want, sorry
12:02:11 Join Bagder_ [0] (
12:02:17linuxstbI should probably just wait until aliask is around, rather than wasting time investigating...
12:02:46 Join tierra [0] (
12:03:00Nico_Plinuxstb: have you looked at the usb log?
12:03:29Dark_ApostropheHas any progress been made when it comes to album art?
12:03:42linuxstbNot yet, no. The first thing I want to do is to work out how aliask is currently running code.
12:03:53linuxstb(that was to Nico_P...)
12:04:06Nico_PDark_Apostrophe: I have something here
12:04:36Nico_Pstill needs some work, but the basics are ready
12:04:43linuxstbNico_P: It seems the merge0.cpp I'm looking it reads the input file 4 bytes at a time, and stops when it reads 0x00000000...
12:04:51Dark_ApostropheNico_P: Great
12:05:04tierrais there a wiki page that says why RBUtil was moved from wxwidgets to qt?...i cant seem to find one
12:05:09Dark_ApostropheNico_P: Have a rough estimate of when it'll be done?
12:05:13Dark_ApostropheA week? A month?
12:05:14Nico_PDark_Apostrophe: no
12:05:22Nico_Pcloser to a week though
12:05:37Dark_ApostropheSounds great :D
12:05:44linuxstbtierra: The reason was that all the devs working on rbutil preferred qt. There were some limitations with wxwidgets, but I forget what.
12:05:47 Quit ackbahr (Read error: 110 (Connection timed out))
12:05:59Dark_ApostropheQt > *
12:05:59 Join ackbahr [0] (
12:06:47 Quit jake_ (Read error: 110 (Connection timed out))
12:06:57 Join jake_ [0] (
12:07:25 Join FOAD_ [0] (
12:07:55 Quit mrkiko ("Lost terminal")
12:08:04tierra(this is actually scorche at a wxwidget dev's house)...
12:08:13tierraNico_P: it doesnt say on that page
12:09:04Nico_Poh I hadn't seen you wanted to know *why*
12:09:23linuxstbtierra: I guess you should ask bluebrother - he started the Qt port...
12:09:24Nico_PIIRC it's just that Domonoky and blueborther were familiar with Qt and not wx
12:09:49Nico_Pand noone was working on the wx version anymore
12:10:11linuxstbDomonoky did a lot of work on the wx version...
12:10:21linuxstbBut yes, he seemed happy to switch to Qt.
12:10:54Nico_Pyeah, I meant "apart from them"
12:11:46linuxstbThat's still the case...
12:12:26 Join Lear [0] (i=chatzill@rockbox/developer/lear)
12:15:33tierralinuxstb: aye...i actually came in here looking for him, but i couldnt see him here...i am out of the state this weekend, and am having chats about random stuff such as wxwidgets with a few friends :)
12:15:41 Join mrkiko [0] (
12:16:19 Join webguest61 [0] (i=55a46843@gateway/web/cgi-irc/
12:16:32webguest61Hi :)
12:16:49n1swhy do we link _every_ plugin against libmad?
12:17:44linuxstbn1s: Because the build system isn't smart enough to allow per-plugin dependencies...
12:18:06linuxstbBut I guess no plugin is using libmad any more...
12:18:10webguest61Can anyone tell me where i can view how much the port for ipod classic and touch is going?
12:18:26linuxstb−−-><−−−− That much.
12:18:36n1slinuxstb: it's fun because the current gccsnapshot I'm testing barfed on libmad so now i can't build any plugins :)
12:19:17linuxstbwebguest61: There aren't any ports for the new ipods, so nothing to see.
12:19:29webguest61in percent
12:19:50 Quit Bagder (Read error: 110 (Connection timed out))
12:20:31 Join lee-qid [0] (
12:21:51webguest61Too bad, is there no activity on that front? After loosing my ipod nano i'm in for a ipod Classic ery soon..
12:22:11linuxstbThat's what I've been trying to say - "no activity"...
12:22:48 Quit FOAD (Read error: 110 (Connection timed out))
12:22:48 Nick FOAD_ is now known as FOAD (
12:23:40Learn1s: there were some fairly big m68k changes a few days ago (like some scheduling support); could be interesting to see what effect they had...
12:24:13n1sLear: yes I noticed but now it won't build libmad...
12:24:39*amiconn doubts that we'll see any significant speedup from using a newer gcc
12:25:16*Lear saw some big faad speedups recently
12:25:17mrkikoamiconn: does your player neither crash with equalizer on?
12:25:20amiconnA few percent, perhaps... unless they added emac support
12:25:25amiconnmrkiko: I never use eq
12:25:59LearThough I don't remember if I actually listened to a file with that build. :) Did run test codec though, and that was some 30% faster.
12:27:13LearTremor was something like 20% slower though...
12:27:21 Quit webguest61 ("CGI:IRC (Ping timeout)")
12:27:44 Join kubiix [0] (
12:28:42mrkikoamiconn: can you try the same experiment with eq on?
12:29:12 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
12:30:06n1sLear: Both Tremor and Mad breaks at above O levels now so benchmarks will not be very useful with those two...
12:30:54LearSounds like we'll have to wait for those changes to stabilize a bit...
12:31:07 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
12:32:04Nico_Pwhat is it that makes code break at too high -O levels?
12:33:06Nico_PI guess my question is really: does it mean the compiler generates bad code by trying to optimise too much? if so, how is that allowed?
12:33:54n1sNico_P: i one case gcc generated invalid asm and in the other it couldn't find registers for a "reload" in/after a block of inline asm that uses lots of registers
12:34:17 Part przemhb
12:34:41linuxstbNico_P: There seems to be a problem with AC3 playback - the codec is going into an infinite loop at the end of the file...
12:35:28linuxstbIt seems ci->request_buffer() succeeds past the end of the file...
12:36:16preglowLear: don't expect them to stabilize if we don't tell them
12:36:28preglowLear: i think the coldfire part of 4.3 won't be touched much more until it's out
12:36:39tierrabluebrother: hi
12:36:43 Join _jhMikeS_ [0] (n=jethead7@rockbox/developer/jhMikeS)
12:36:43 Quit jhMikeS (Nick collision from services.)
12:36:58Nico_Plinuxstb: sounds a bit like the bug I had with MP3 and thought I had fixed :/
12:37:36 Quit spiorf (Read error: 110 (Connection timed out))
12:37:53tierrabluebrother: i was wondering why you made the switch to QT from wxwigdets for RButil (this is scorche over at a wxwidget dev's house and i was at a loss, and searching fails me)
12:37:57 Join spiorf [0] (
12:38:24linuxstbNico_P: The main decoding loop in a52.c is very simple - it just requests 4096 bytes with ci->request_buffer(), calls the a52 decoder to decode that data, then advances the buffer 4096 bytes. When ci->request_buffer() returns zero, it stops.
12:38:39linuxstbI mean sets the first parameter to zero
12:38:46 Join Bubelgum [0] (
12:39:13n1sooh, panic :)
12:39:15Nico_Plinuxstb: the filesize is a multiple of 4096?
12:39:26linuxstbNot always.
12:39:42Dark_ApostropheHave any of you ever played SuperTux?
12:40:09Dark_ApostropheIt's an open source side-scrolling supermario-type game... Would that theoretically be portable to Rockbox?
12:40:21stripwaxtierra - I think internationalisation/localisation was deemed easier with Qt than with Wx - and/or accessibility (screen readers)
12:40:22 Nick bb_ is now known as bb (n=bb@unaffiliated/bb)
12:40:33tierraDark_Apostrophe: feel free to port it yourself
12:41:11Nico_Phmm I need an AC3 file
12:41:19tierrastripwax: from what i understand, wxwidgets has a bit better accessibility due to being just a wrapper over the OS's api
12:41:40Dark_Apostrophetierra: I'll have to learn to code first
12:41:56tierraDark_Apostrophe: never a better time to learn :)
12:42:10Dark_ApostropheWhat language is Rockbox written in?
12:42:19markunNico_P: I PMed a link to a AC3 file
12:42:36Dark_Apostrophetierra: Hrm...
12:43:28bluebrothertierra: hi :)
12:43:49 Nick bluebrother is now known as blue|coffee (n=dom@rockbox/staff/bluebrother)
12:43:54Dark_Apostrophetierra: I gave up on learning Python, which is far easier, even. :/
12:44:25markunNico_P: just seek to the end
12:44:37n1sDark_Apostrophe: but now you have motivation :-)
12:44:40Nico_Pmarkun: is it normal that I hear nothing?
12:44:45Nico_P(sim here)
12:44:49blue|coffeetierra: well, various reasons. Most of all, I like Qt much better and wanted to know if I could port it ;-)
12:44:55Dark_Apostrophen1s: Not really, as I play SuperMario on rockboy :p
12:44:59blue|coffeeand wx looks rather horrible on my linux box.
12:45:08Dark_Apostrophen1s: But I just find SuperTux to be somewhat better than mario
12:45:24blue|coffeealso, there were issues with static linking on linux prior to the switch
12:45:30markunNico_P: very silent :)
12:45:36Nico_Poh yes
12:45:44 Join theclaw [0] (n=theclaw@unaffiliated/theclaw)
12:45:45markunbeginning of Blade Runner
12:45:47n1sI guess having super tux ported will make more users think we are linux based...
12:45:57tierrablue|coffee: do you know the technical reasons?
12:46:05linuxstbtierra: Ah yes, the Linux distros seemed to be fed up supporting all the different incompatible versions of libwx... e.g. Debian didn't seem to be incorporating 2.8 any time soon...
12:46:09Dark_Apostrophen1s: No more than frozen bubble, which is also a Linux game
12:46:44Dark_Apostrophewx = nasty
12:46:45linuxstbYou mean it's a game also available for Linux?
12:46:48n1syeah but tux is somewhat more associated with linux :)
12:46:59 Join mf0102 [0] (n=michi@
12:47:11blue|coffeeon linux wx uses gtk. I think that was a reason for the linking issues but I never investigated them
12:47:23theclawWhile playing a relatively long mp3 file (~30 min), rockbox crashed on my sandisk sansa e280. Did something like this happen to anyone else here? (I have to know, as it might as well be an hardware error)
12:47:52blue|coffeeinitially it was an interesting challenge for me as I hadn't done much cpp before
12:48:10*tierra transfers to the wxwidgets dev for a sec out of interest
12:48:13markunNico_P: do you get the infinite loop at the end as well?
12:48:14blue|coffeeand people quite liked it ... IIRC the idea of using Qt came up much earlier
12:48:21Nico_Pmarkun: no
12:48:28Nico_Pmarkun: do you have it on the sim?
12:48:45markunthe only thing I can do is kill -9
12:48:54Dark_ApostropheSuperTux is written in C++. Hrm. :/
12:49:02markunlinuxstb has the same problem
12:49:14Dark_Apostrophen1s: Err...
12:49:19Nico_Phmm I wasn't very up to date
12:49:20markunNico_P: I'll svn up, is your build up to date?
12:49:22Dark_Apostrophen1s: Ever played frozen bubble?
12:49:25Nico_Ptrying again
12:49:28Dark_Apostrophen1s: All you see are penguins
12:49:41tierralinuxstb: that is true, we lost our Debian maintainer due to some ugly arguments between other devs, so we really have no idea when those packages will be picked back up unless the Ubuntu packages are used
12:50:26Nico_Pstill no loop
12:50:29tierraon that note, Gentoo still doesn't have 2.8 either due to some issues with portage (I'm not entirely aware of what's involved there)
12:51:25Nico_Pmarkun, linuxstb: I fixed something very similar in r155'1
12:51:32Nico_Psorry, r15541
12:52:15 Join moos [0] (
12:52:21*linuxstb updates
12:52:56*markun does the same
12:53:02 Join karashata [0] (
12:53:03linuxstbNico_P: Sounds like a false alarm then...
12:53:39Nico_Phaha, lostlogic and I always ask people who report bugs here to update and retest... this time I forgot :p
12:54:00*blue|coffee still has no internet at home :'-(
12:54:06markunlinuxstb: would be very bad timing if we both updated during that 3 hour window
12:54:32Nico_Pmarkun: do you know what rev you had?
12:54:39linuxstbblue|coffee: And no friendly neighbours with open wifi?
12:54:45theclawwouldn't it be better to port SDL to rockbox instead of porting each SDL game?
12:54:56blue|coffeeunfortunately not.
12:54:57theclaw(would it be possible?)
12:55:02markunNico_P: not anymore..
12:55:23Nico_Pmarkun, linuxstb the fix was on Thursday... surely you wer more up to date than that?
12:56:02linuxstbYes, fixed now...
12:56:10markunhere too
12:56:14*markun feels stupid
12:56:16linuxstbI guess not...
12:56:23*mrkiko feels tired
12:56:26 Nick linuxstb is now known as stupid (n=linuxstb@rockbox/developer/linuxstb)
12:57:15markunstupid: thanks for debugging anyway
12:57:40*stupid is always available
12:57:56 Nick stupid is now known as linuxstb (
12:58:06mrkikoI have an idea for "blind debugging".
12:58:34markunlinuxstb: not sure what to do with my ambisonics code. Right now it just decodes to stereo.
12:58:52linuxstbIs it currently part of the wav codec?
12:59:01mrkikoWouldn't be possible to put rockbox in a special mode where a log file is written to disk synchronously? This way it would be possible to trigger crashes and observe then what happened. The problems are various - the one I see (and not the bigger) is the fs may become corrupted badly.
12:59:16Nico_Plinuxstb: I have a question: is the cost of opening and closing a file two or three times big enough that I should bother to try reducing the number of opens?
12:59:50markunlinuxstb: yes
13:00:43linuxstbmarkun: I would say go ahead and commit it...
13:00:53markunthe code should simulate 2 microphones pointed at 45 and -45 degrees to the front
13:01:06linuxstbAlthough your assumption about multi-channel audio seems wrong...
13:01:08markunlinuxstb: I'll optimize the code a little bit then
13:01:26markunlinuxstb: assuming that it's always ambisonics you mean?
13:01:26linuxstbCan't you check for that ambisonics flag?
13:01:40markunyes, I will do that
13:02:39mrkikooh... I should escape!
13:02:40 Quit mrkiko ("leaving")
13:03:32linuxstbNico_P: I'm not sure. It seems inefficient to do multiple opens and closes though, if they can be avoided...
13:04:23Nico_Pyeah I agree
13:04:48*Nico_P seems to recall we used to read_bmp_fd in addition to read_bmp_file...
13:04:50 Part daurnimator
13:05:01Nico_P"used to have", that is
13:05:14***Saving seen data "./dancer.seen"
13:06:45preglowmarkun: that really doesn't simulate speakers at -45 and 45 degress in any way you can justify, though. like i said, those equations are only accurate if you decode to enough speakers
13:07:11markunpreglow: alright
13:07:38preglowit might sound as if the sound comes from two speakers in front of you, but that's more because of the crossfeed, i'd imagine
13:07:49 Join Shike [0] (i=5966cf08@gateway/web/cgi-irc/
13:07:53markunpreglow: yes, of course
13:08:40markunbut for the crossfeed to make sense I should find a way to simulate microphones pointing at those directions
13:08:47preglowi guess so
13:10:03markunI didn't find any equations to generate any posision for cardioid and figure-of-eight microphones yes, only some examples
13:10:57markunpreglow: do you have them at hand?
13:11:11preglowdidn't wikipedia have a couple?
13:11:28markunonly examples
13:11:48preglowwell, why can't those be used?
13:12:17 Quit Shike (Client Quit)
13:12:39markunnon of them match what I want to do, I want cartioids pointing forward at 45 and -45 degrees
13:13:45preglowwell, if i had a proper mathmematical background on spherical harmonics, i might be able to help you :)
13:14:45markunI thought you did
13:15:15preglowi'm a bit shady on much of the lower level math stuff going on
13:15:33markunpreglow: btw, didn't you tell me about the bass problem at very high orders?
13:15:36 Quit BigBambi (Remote closed the connection)
13:16:32markunpreglow: did you see this?
13:16:49amiconnpreglow: Do you have an idea how efficient speex is now on PP, especially PP5002?
13:17:01 Join daurnimator_ [0] (n=daurnima@
13:17:35preglowmarkun: both seen and implemented
13:17:47preglowamiconn: it's ok on my pp target, i have no idea how good it is on pp5002
13:18:03preglowamiconn: sample files here, if you want to try
13:18:17preglowamiconn: no code is in iram yet
13:18:20Nico_Pany objections about ? amiconn?
13:19:06n1sLear: something is very broken, my h300 either pancics right after booting or shuts down immediately with a low battery warning (even with the patch to the pcf driver that used to fix it)
13:19:10preglowamiconn: nb = 8khz, wb = 16 khz, uwb = 32khz, _vbr files are better quality
13:19:43preglowneed a shower, brb
13:20:46amiconnNico_P: What's that good for?
13:20:47 Part MouFou ("Kopete 0.12.6 :")
13:21:04Nico_Pamiconn: if I already have an fd to the bitmap
13:21:12Nico_Pso reducing open() calls
13:22:54Nico_Pamiconn: is that worth it? if not I'm happy to continue using read_bmp_file
13:29:30Nico_Pamiconn: thoughts?
13:30:12 Join UncleRemus [0] (
13:32:04UncleRemusJust put rockbox on my iPod (Photo, 4th G) fr the first time! How do oneself add songs to it? Just copying over to a folder, or what? (I'm running Linux and got it mounted at /media/ipod. Any nice clients?)
13:32:09 Join blue|cof1ee [0] (
13:32:25n1sUncleRemus: just copying works fine
13:33:21stripwaxUncleRemus - don't need a client
13:33:54UncleRemusOK, just saw that amarok supports it somehow.
13:33:54stripwaxUncleRemus - you might want to create a top level Music directory but other than that you can copy the music wherever you like :)
13:35:14UncleRemusThanks guys. What do you say about mp3s vs. flacs on it? Will flacs cause it to run out of battery faster?
13:35:27n1sUncleRemus: yes
13:35:36stripwaxNico_P - any chance you'd be looking to update my sliding_puzzle patch as part of the albumart thing ? :)
13:35:48stripwaxor even committing it or something? :)
13:35:56UncleRemusThat's OK for me. iPod is not a hifi-experience for me anyhow.
13:36:13preglowmarkun: it can synthesize any response
13:36:16stripwaxUncleRemus - there's a really good table of codec vs battery life on the wiki
13:36:22Nico_Pstripwax: after the first commit, yeah, if I don't forget ;)
13:36:25stripwaxIpodRuntime i think
13:36:29stripwaxNico_P : ;-)
13:36:36amiconnn1s: uh?
13:36:50n1samiconn: ?
13:36:55UncleRemusstripwax: found it. looks nice!
13:37:02Nico_Pstripwax: how does it get the AA?
13:37:16amiconnI'd expect longer battery runtime with flac than with mp3 on ipods, especially pp5020 ipods
13:37:27Nico_Pamiconn: any thoughts about my question above?
13:37:35stripwaxNico_P - currently? I was asked by rockboxdevs to copy+paste the relevant bits of the albumart patch and put it into the sliding_puzzle patch ...
13:37:38markunpreglow: here's some info about it:
13:37:38Nico_P(sorry to insist)
13:37:38stripwaxalways happy to oblige
13:37:40amiconnI wonder what you're trying to do...
13:37:51Nico_Pamiconn: me?
13:37:54stripwaxbut albumart is only a part of the sliding_puzzle patch
13:38:15Nico_Pamiconn: trying to reduce the number of open() calls by reusing an existing fd
13:38:27stripwax(so I guess I'm still not sure why it's not been committed yet, but .. ) having albumart committed should be the perfect excuse to have it dusted off
13:38:37amiconnYeah, but where/what for??
13:38:53Nico_Pamiconn: it's meant for AA, which I'm working on
13:39:09blue|cof1eeok, in case nobody was able building / testing: I put a current svn rbutil build at
13:39:16Nico_Pin my code I have the fd but call read_bmp_file, which is a bit frustrating
13:40:38linuxstbDidn't the tar-for-wps patch include a change to read_bmp_file so it could take an fd?
13:40:55preglowthat patch needs some stroking and inclusion, btw...
13:41:04Nico_Plinuxstb: ah, that might be where I saw the idea
13:41:13n1samiconn: ah, according to the IpodRuntime wiki page they seem to give about the same runtime but there are no directly comparing tests on the same device asaict and all flac tests are from last year...
13:41:14linuxstbpreglow: When Nico_P and I tested it, we saw no noticable speedup...
13:41:41amiconnWell, if you think that's necessary...
13:41:53n1sUncleRemus: sorry you will probably see about the same runtime with flac and mp3
13:41:56*amiconn still doesn't understand why AA seems so important for many people
13:42:05amiconnAnyway, gotta reboot
13:42:08 Quit amiconn (" HydraIRC -> <- Now with extra fish!")
13:42:16Nico_Pamiconn: well the question boils down to "is the cost of open() big enough to warrant the change?"
13:42:22preglowlinuxstb: then what makes wps loading so slow?
13:42:30Nico_Ppreglow: it's slow?
13:42:37linuxstbpreglow: I don't find it that slow...
13:42:40UncleRemusn1s: Yeah, looked like that
13:43:00preglowi thought we all agreed it was slow :V
13:43:02 Quit daurnimator_ ("Cyas")
13:43:14linuxstbI think that could be the problem - it may depend on how fragmented the disk is, and Nico_P and I have clean disks... That patch needs to be tested by people experiencing slow loading...
13:43:18UncleRemusI will now remove everything on my iPod and only have .rockbox directory left, this is OK, or?
13:43:33preglowlinuxstb: if you ask me, it's rather obvious it'll be faster in _some_ way
13:43:34UncleRemusI'm leaving iTunes and gnupod
13:43:38 Join daurnimator_ [0] (n=daurnima@
13:43:41preglowhaving a zillion bmps around isn't very pretty
13:44:04linuxstbpreglow: Yes, but it adds complication - so we didn't think it was worthwhile unless it was shown to give a noticable speed improvement.
13:44:30n1sUncleRemus: yep, it's a-ok :)
13:44:31UncleRemusI hope I don't screw up my rockbox then, am I?
13:45:04UncleRemusI'm really excited about this. Can't believe I haven't found it earlier
13:45:32preglowgrouped bmps should go in the same bmp as well
13:45:38preglowgrf, so many things i want changed, so little time...
13:46:04 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
13:46:32 Quit blue|coffee (Read error: 113 (No route to host))
13:47:57 Join Arathis [0] (
13:48:03preglowamiconn: i'd be very interested in performance figures if you do test, btw
13:49:36 Join mirak [0] (
13:50:29amiconnAh, that
13:51:10amiconnThat's why I needed to reboot - windows insisted that the SBP2 device has a probelm that only a reboot could fix, even if it worked without that huhdreds of times before
13:51:10 Quit kubiix (Read error: 104 (Connection reset by peer))
13:51:22jmspeexpreglow: how's stereo.c doing?
13:52:02amiconnpreglow: You made a whole bunch of test files - is there a table somewhere showing your cf and pp5020 results?
13:52:17amiconnAlso, what are the most interesting tests
13:52:27*amiconn won't be able to test all of them right now
13:52:55preglowjmspeex: oh, fine, haven't done any more with it today
13:53:00preglowamiconn: nope
13:53:13preglowamiconn: all of them aren't needed, i'd say test the q8_vbr of each
13:53:30preglowq4 will be slightly faster, but not much
13:53:57jmspeexpreglow: BTW, about having raw files, I've got another suggestion. You could pad bytes for each packet end encode a 1-char length. That would waste on average about 600 bps, but would make it easy to convert these files to/from Ogg.
13:54:21preglowjmspeex: i doubt anyone would want to convert them to ogg
13:54:35preglow600 bits per second isn't much, though :)
13:54:39 Join DerPapst [0] (
13:54:52jmspeexpreglow: well, with the space saved by the ogg library, you could put the encoder in :-)
13:55:30preglowi will try porting the encoder one day, but that'll be its own thing, it won't go in the decoder plugin anyway
13:56:37preglowand then i probably won't bother disabling the decoder, the main reason i spent time disabling the encoder is it's going in the core
13:57:31jmspeexpreglow: well, might as well put the encoder in the core as well :-)
13:57:44amiconnPP5002 results: nb q8 vbr: 501%, wb q8 vbr: 278%, uwb q8 vbr: 169%
14:02:10UncleRemusNow it says Scanning disk... and I can't do nothing, I have only a few MB's on it currently. Should I await this process?
14:03:08preglowamiconn: will check and compare
14:03:11preglowwith pp5020
14:03:41amiconnThat means uwb speex is slower than mp3 @128kbps - even though that is stereo and 44.1kHz
14:03:56amiconnalmost exactly 200% on PP5002 for the latter...
14:04:03preglowputting a couple of functions in iram will help alot
14:04:10preglowlots of time is spent in iir_mem16 and qmf_synth
14:04:12amiconnYes, most probably
14:04:24preglowwb q8 vbr: 438%
14:04:26preglowjust for comparison
14:04:55jmspeexamiconn, preglow: I would recommend against using uwb for most things
14:04:58amiconnThe core itself runs at the same speed, so code running from iram working on iram data sees no difference between PP5002 and PP502x
14:05:07preglowjmspeex: we won't
14:05:30preglowjmspeex: btw, do you know what mode/bitrate the sam & max games use?
14:05:37 Part Dave2 ("Pruning my channels.")
14:05:58amiconnBut as soon as sdram, and hence the cache, is involved, speed can drop to 50%
14:06:00preglowamiconn: uqb q8 vbr: 256%
14:06:04jmspeexpreglow: what uses?
14:06:39preglowjmspeex: what speex mode/bitrate those games use
14:06:43preglowi just assumed you know they use speex...
14:07:02jmspeexwhat games?
14:07:10preglowthe newly release sam & max games
14:07:18preglowmade by telltale games
14:07:19jmspeexwhat's that?
14:07:38Nico_Pdoes someone know if there is a point, to determine if a file exists, in trying to use dircache before trying to open it?
14:07:56 Join GnosticE [0] (
14:08:04preglowjmspeex: just tried a free episode i got of it yesterday, and i noticed they use speex
14:08:16preglowi've developed a pretty good ear for detecting speex files, for some reason...
14:08:16 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
14:08:21jmspeexhow did you notice?
14:08:22amiconnuwb q4 vbr is 175%
14:08:54preglowjmspeex: characteristic sibilance distortion
14:10:01preglowjmspeex: btw, why don't you recommend using uwb? too little gain for the extra file size?
14:10:04jmspeexpreglow: care to describe?
14:10:31preglowjmspeex: it's kinda hard to explain
14:10:48amiconnA quick listening test makes me think that wb q4 vbr would be good enough for voice files
14:10:55preglowi just recognize it after having listened to a ton of speex encodings of the same file :)
14:11:06preglowamiconn: i was planning on using q5-6
14:11:17amiconnThere are no q2 and q3 test files
14:11:19jmspeexpreglow: extra size is small. The problem is that it's sometimes too small and the added distortion can make things sound worse to some people.
14:11:38preglowamiconn: there are however q5-7 files there for wb
14:11:42jmspeexuwb too was a one-evening hack
14:11:45preglowamiconn: i can encode more if you want
14:12:11preglowjmspeex: i pretty much decided on using wb anyway, i don't see much point in uwb for speech
14:12:34amiconnSkipping back doesn't werk - looks like a MoB bug
14:12:34 Quit einhirn (Read error: 104 (Connection reset by peer))
14:12:44SlasheriNico_P: yes, using dircache is very much faster if that is available
14:12:49jmspeexpreglow: also note that other CELP codecs would tend to sound at least a bit similar. Though they're most likely using Speex for cost reasons :-)
14:13:24preglowjmspeex: i pretty much know they do since i asked a guy who had checked it out
14:13:30markunjmspeex: nice feeling to see your baby ending up in all kinds of products? :)
14:14:30preglowamiconn: q2 and q3 uploaded now
14:14:38*amiconn wonders how speex will cope with synthesized voices
14:14:41jmspeexmarkun: it's nice to see it in games, but what I'm really after is seeing it in every IP phone.
14:15:56 Join einhirn [0] (
14:17:00preglowjmspeex: btw, i shouldn't need clipping in the stereo decoder, should i? the ouput should always be within bounds?
14:17:18 Quit daurnimator_ ("Cyas")
14:17:29jmspeexamiconn: should cope with synthetic voice fine I'd say. I'll cope with anything periodic.
14:18:31 Nick Bagder_ is now known as Bagder (
14:18:58amiconnA quick check makes me think that even wb q3 should be enough for voice, and better than what we have now, at lower bitrate
14:19:18amiconnOf course that's only possible for swcodec
14:19:31preglowunless someone does a mas codec...
14:19:49preglowmas would be very capable of decoding speex, for sure
14:19:51amiconnBut would that even be possible?
14:20:08amiconnMany newer codecs need to many/too large data structures to run them on the mas
14:20:11preglowspeex has both small data and code and does very typical dsp things
14:20:34preglowso yeah, my guess is mas could do it
14:20:35amiconnThe MAS core has 4K words of RAM
14:20:43Nico_PSlasheri: but I see open() already tries to use it, so for a file_exists() func, should I just use open or is there an advantage in using dircache_get_entry_ptr()?
14:20:44preglowbut it's not really a possibility anyway...
14:20:45amiconnThat needs to hold both code & data
14:21:01preglowamiconn: what is a mas word? 16 bits?
14:21:02Nico_PSlasheri: knowing that I don't need the FD, of course
14:21:02amiconn...and transfer buffers from the host (although these can be small for speex)
14:21:05amiconn20 bits
14:21:43 Join Crash91 [0] (n=evil91@
14:21:49amiconnBut yes, processing power should be sufficient for speex
14:21:58preglowfor both code _and_ data
14:22:05amiconnThe core runs at 22..24 MHz
14:22:07preglowi don't know how they manage to put mpeg in there, then...
14:22:24amiconnThe mpeg codec is in ROM, so they need the RAM only for data
14:24:28Crash91is there any way to customize the frequency range the rockbox radio listens to? (i want to be able to hear airband on my sansa)
14:25:39 Join stewball`ghost [0] (n=WTFOMGBB@
14:25:54amiconnThe MAS3587F contains mpeg audio layer 2 and 3 decoder, and layer 3 encoder. The MAS35x9F family contains mpeg audio layer 2 and 3 decoder, G.729 decoder, and AAC decoder, in various combinations
14:26:30amiconnOf that family, only the MAS3539F is used in a rockbox target, and the 3539 is mpeg layer 2+3 only
14:26:46SlasheriNico_P: hmm, there is still small advantage in using direct dircache call as that doesn't have to access disk at all
14:27:01amiconnOnly the ROM content is different for all mentioned MAS', the core, interfaces and RAM is all the same
14:28:32Nico_PSlasheri: ok, I'll try thats then
14:28:39 Join desowin [0] (
14:30:28jmspeexamiconn, preglow: what's MAS?
14:31:02preglowjmspeex: the hardware decoder chip in our older archos targets
14:31:10preglowjmspeex: it's a combined custom dsp core and dac/adc
14:31:26jmspeexFYI, Speex runs duplex on a dsPIC, which has 8 kB RAM
14:31:39preglowboth encoding and decoding?
14:31:50jmspeexnarrowband though
14:31:53preglowdoes that ram include const data?
14:32:09 Part pixelma
14:32:13preglowthat'd probably be the problem, speex needs more const data than it needs ram
14:32:20preglowat least non-stack ram
14:32:53jmspeexpreglow: The decoder needs about 1.5 kB RAM for the state and another 1.5 kB for stack
14:33:22preglowso wb mode is what makes the stack usage rise to about 4k?
14:33:22jmspeexif you only pick one mode, you can probably get away with only 2-4 kB worth of codebooks
14:34:06jmspeexwideband would about double that
14:34:07preglowstill, that's not bad, what clock rate do those run at?
14:34:10Crash91is there any way to customize the frequency range the rockbox radio listens to? (i want to be able to hear airband on my sansa)
14:34:32preglowCrash91: if it's not included in any of the usual region settings, you'd need to modify the source
14:35:08jmspeexpreglow: no idea what clock rate. I've never used these chips and the port is actually proprietary (made by the company from an old 1.1.x version)
14:35:27preglowright, no source code?
14:36:03 Quit blue|cof1ee (Read error: 113 (No route to host))
14:36:53 Quit mf0102 ("Verlassend")
14:36:59jmspeexno, just a "free demo". Kind of stupid for a chip manufacturer.
14:37:57preglowi would say
14:39:41jmspeexOn the opposite end, I've had some contracts from Analog Devices to optimise Speex for their Blackfin chip. That's overall the best port so far I'd say.
14:39:41preglowbbl, shop
14:39:55preglowahh, yeah
14:39:58jmspeexThen there's TI that doesn't seem to care much
14:40:07preglowthat was part of the reason you went 16 bit, wasn't it?
14:40:19preglowti have never been very nice people...
14:40:46preglowbut yeah, bbl
14:40:47jmspeexpreglow: the move to 16-bit was mainly "because I realised I could".
14:41:33jmspeexI had to teach myself how to do fixed-point when converting Speex.
14:41:34theclawis it possible to debug the rockbox source directly on the player? (sansa e200)
14:42:21n1stheclaw: depends on what you mean
14:43:55theclawn1s: stepping through the code and seeing the results on the player ;)
14:44:48n1stheclaw: you'd need jtag for that, I believe it's possible though (with some soldering etc)
14:44:50 Nick fxb is now known as fxb__ (
14:47:53amiconnjmspeex: For speex support on MAS, the codec, plus data, plus input buffer, plus code to control the audio output would need to fit into 4K words (that's 10K byte)
14:47:58 Part Crash91
14:48:25amiconnThe core is a custom dsp (harvard architecture). There's no public docs about the core, but some leaked ones, plus a leaked simulator
14:48:57 Join Robin0800 [0] (
14:49:59*amiconn disappears
14:51:36 Join webguest99 [0] (i=50828ab8@gateway/web/cgi-irc/
14:55:15webguest99hi! I've got a question(maybe it's been asked once before, but I couldn't get any pieces of information about that^^): How high do you estimate the risk of damaging my sansa if I install rockbox on it?
14:56:38 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
14:57:06 Quit markun (Remote closed the connection)
14:57:41n1swebguest99: if you follow the instructions the risk is pretty much 0
14:57:48jmspeexamiconn: 8 kbps mode might fit in about 4 kB, plus the code size. I don't know how compact the instructions are
15:00:28jmspeexactually, if you disable the enhancer, then the decoder becomes really really small and would easily fit on that chip. Encoder is less sure.
15:00:35 Quit webguest99 ("CGI:IRC (Ping timeout)")
15:05:17***Saving seen data "./dancer.seen"
15:05:53 Join markun [0] (n=markun@rockbox/developer/markun)
15:08:21Nico_Pis a HAVE_ALBUMART #define going to be wanted?
15:09:59 Join stripwax [0] (
15:10:28DerPapsti bet yes ;)
15:11:25DerPapstfor people who don't want it to keep binsize small and boottime low ;)
15:11:29JdGordonNico_P: ptobably not.. only add the code on bitmap and your fine
15:11:40JdGordonAA in MoB?
15:12:05Nico_Pit's working since yesterday evening
15:13:44n1sI think adding in those defines are a bad Idea unless it helps with debugging, people who want to exclude features can change the code themselves
15:15:54linuxstbDo we want album-art on mono LCDs?
15:16:00markunn1s: why do you think it's a bad idea?
15:17:02markunlinuxstb: and on charcel LCDs :)
15:17:14n1smarkun: clutters the code with useless #ifdefs _and_ those ifdefs are likely to become broken if every change to the specific feature or something relating to it is not built with them #undefed...
15:17:50markunwell, I don't agree that the ifdefs are useless
15:18:18n1sbut if it is only in a couple of places i don't mind really I just think we should keep it to a minimum
15:21:01Nico_Psome people might want to disable the feature in their builds... and currently I have a lot of #ifdef HAVE_LCD_BITMAP, which would all move to #ifdef HAVE_ALBUMART, allowing us to control better which targets get AA
15:21:28preglowjmspeex: but enhancing is pretty essential for celp codecs in general, yes?
15:21:32Nico_Por I change #ifdef HAVE_LCD_BITMAP to #ifdef HAVE_LCD_COLOR
15:22:45Nico_PI also need to decide what to do with all the align and resize flags
15:23:22n1sNico_P: are you going to let the wps dictate the size of the image or is it fixed or do you scale?
15:24:41Nico_Pn1s: currently there is no scaling
15:25:07n1sbut you allow different sizes?
15:25:11Nico_Pso the WPS fixes the size of display, but the bitmaps are loaded at their original size
15:25:34n1sso what dou you do if the image doesn't fit?
15:26:06Nico_PI think it's cropped... not completely sure as I haven't done much testing of that part yet
15:27:09Nico_Pbut it's been in the patch for quite long so I assume it's been tested
15:27:21preglowmerbanan: did anything fun happen to aac yesterday?
15:29:16 Join sarixe [0] (
15:30:27 Quit JdGordon ("Konversation terminated!")
15:34:58preglowamiconn: btw, did you try to put iir_mem16() and qmf_synth() in iram? they're both in filters_arm4.S
15:46:42 Quit ackbahr (Remote closed the connection)
15:47:57 Quit karashata (Nick collision from services.)
15:48:19 Join karashata [0] (
15:50:59 Quit Arathis ("Bye, bye")
15:53:40Nico_Pn1s: I've changed it so that bitmaps that are too big are cropped
15:53:54Nico_PI plan on including scaling, but later
15:57:46 Quit donutman25 ("ChatZilla [Firefox]")
15:58:23 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
15:59:05 Join mrkiko [0] (
15:59:09mrkikoHi all!
15:59:45 Quit desowin (Remote closed the connection)
15:59:59 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
16:00:10 Quit daurnimator (Read error: 104 (Connection reset by peer))
16:00:36 Join desowin [0] (
16:04:39 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
16:04:52 Quit daurnimator (Read error: 104 (Connection reset by peer))
16:04:54 Quit einhirn (Read error: 104 (Connection reset by peer))
16:07:30 Join webguest76 [0] (i=54ac5fce@gateway/web/cgi-irc/
16:08:07 Part webguest76
16:15:26 Quit Toxicity999 (Remote closed the connection)
16:17:32 Quit daurn (Read error: 104 (Connection reset by peer))
16:18:02 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
16:18:28 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
16:18:41 Join ackbahr [0] (
16:19:10 Quit daurnimator (Remote closed the connection)
16:20:30 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
16:21:02 Quit Lear ("ChatZilla [Firefox]")
16:22:23 Quit Robin0800 (" Want to be different? Try HydraIRC -> <-")
16:22:39 Join Robin0800 [0] (
16:22:40*freqmod_nx has a speex file that only works if you play another one first, (if not you get silence)
16:23:40 Quit Robin0800 (Client Quit)
16:23:55 Join Robin0800 [0] (
16:24:40 Quit ackbahr (Remote closed the connection)
16:25:46preglowfreqmod_nx: i know about the problem, will fix when i find out why
16:26:23freqmod_nxno problem, I just thought I should report it
16:26:45preglowfreqmod_nx: quick thing you can try: in nb_decoder_init, memset the entire static struct to 0
16:26:55preglowfreqmod_nx: also in sb_decoder_init if it's a wb file
16:29:27 Join Crash91 [0] (n=evil91@
16:29:28freqmod_nxpregi will try it after i have made dinner
16:29:52freqmod_nx*preglow: (stupid tab completion)
16:29:54 Quit Bubelgum (Remote closed the connection)
16:35:01 Quit Toxicity999 (Remote closed the connection)
16:36:00*n1s spots // comments in speex.c
16:37:04 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
16:37:45 Join ackbahr [0] (
16:41:33Nico_PI think AA is almost ready to commit
16:41:41Nico_Panyone want to test?
16:41:46*preglow is so used to them :/
16:41:49preglowNico_P: how does it work so far?
16:41:50Crash91AA? Album art?
16:42:22Nico_Ppreglow: pretty well. it seems to show a few flaws in the buffering code but most of the time works fine
16:42:25Nico_PCrash91: yes
16:43:29alienbiker99Nico_P AA patch thats been around for ages?
16:43:34Nico_Ppreglow: I get the occasional whitenoise instead of the bitmap
16:43:49Nico_Palienbiker99: I've heavily modified it to adapt it to MoB
16:43:57Nico_Pwhich means it could be committed
16:44:24Crash91Nico_P: what is stopping you?
16:44:24preglowNico_P: well, it's fine by me
16:46:34 Quit stewball`ghost ()
16:46:45 Quit sarixe ("(EE) Failed to load "quit" module")
16:46:49 Join stewball`ghost [0] (n=WTFOMGBB@
16:47:49alienbiker99i might be able to tes the partch. where is it?
16:48:19 Join PaulJam [0] (
16:48:21alienbiker99i cant type this morning =\
16:48:22Nico_Palienbiker99: I still need to pastebin it somewhere but I want to test it on my gigabeat before
16:49:08 Quit billenium (Read error: 104 (Connection reset by peer))
16:50:14Robin0800Nico_P,I can test on an ipod video
16:50:32 Join billenium [0] (
16:51:07freqmod_nxn1s: (//) my fault :S
16:51:08alienbiker99i have a sansa, and h300 right now
16:52:04alienbiker99oh i cant say just sansa anymore. i have an E200
16:53:38n1sfreqmod_nx: I only saw the one preglow added (and now removed)
16:54:56freqmod_nxn1s: ok, I saw one in the seeking code which I don't think is modyfied after my original patch.
16:55:18billeniumi fail at xubuntu, brb
16:55:23 Quit billenium (Remote closed the connection)
16:55:26preglowfreqmod_nx: i'll remove some floats from the seeking code
16:55:41preglowthere's both a floating point mul and div in speex.c, without those, there is no more floats left in our decoder
16:55:59freqmod_nxfeel free, the seeking code is not very good, it's relativly slow, and sometimes it doesn't work
16:56:42freqmod_nxi think i remember that it workes better in the simulator than on target.
16:56:46Nico_Palienbiker99, Robin0800:
16:57:04Nico_Phmm actually I forgot to svn add some files
16:57:27preglowfreqmod_nx: how's that possible?
16:57:32 Quit daurnimator (Remote closed the connection)
16:58:00freqmod_nxpreglow: I don't know, and I din't test it very closely.
16:58:18 Part Crash91
16:58:28preglowbut yeah, we'll see, the primary reason i'm working on speex is to make it a good voice ui codec, and we won't need seeking there
16:58:38freqmod_nxpreglow: was it the st DecState you suggested to memset (in nb_decoder_init?)
16:58:45preglowfreqmod_nx: yes
16:59:07preglowfreqmod_nx: just put one memset call before anything else happens
16:59:29freqmod_nxpreglow: it works mostly, and resume works well. (which is most important for audiobooks etc.)
16:59:40*freqmod_nx looks up the memset api
17:00:45 Quit ackbahr (Remote closed the connection)
17:00:47freqmod_nxpreglow: if the seeking failes, then it will play back from where it were before the seeking.
17:02:37markunscorche: look, we are on this page :)
17:03:09alienbiker99Nico_P. do i just save this as one file?
17:04:11Nico_Palienbiker99: it's the patch, but I need to send a new one
17:04:42preglowfreqmod_nx: any results?
17:05:09Nico_Palienbiker99, Robin0800:
17:05:20Nico_Palienbiker99: you apply the patch to your source tree and build
17:05:22***Saving seen data "./dancer.seen"
17:05:32freqmod_nxpreglow: it's just like before
17:05:51preglowwell, shit
17:05:58preglowcould you send me the file in some way?
17:08:39 Quit FOAD ("I'll be back")
17:11:51 Join FOAD [0] (
17:11:58*preglow vanishes
17:12:45 Quit mirak ("Ex-Chat")
17:13:42Nico_Phmm corrupted bitmap
17:13:49alienbiker99i just got that
17:14:35Nico_Palienbiker99: I think those are problems in the buffering code, but I'm not 100% sure... I'd need to be able to repro them
17:15:22alienbiker99ok. should i remove this patch then?
17:15:59Nico_Palienbiker99: the patch doesn't add any problems (I think), and it certainly isn't putting your player at risk
17:16:24Nico_Pwhat I meant was that these corruptions were indicators of other preexisting problems
17:18:14 Join random_desu_is_s [0] (
17:18:41 Join toffe82 [0] (
17:20:21Nico_Phmm actually I think I know what the problem is
17:24:56 Quit idnar (Nick collision from services.)
17:24:59 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
17:25:36alienbiker99easy to fix Nico_P?
17:25:56Nico_Palienbiker99: yeah, pretty easy, but I need to make sure I'm right
17:26:53 Join Mouser_X [0] (i=cf9bb003@gateway/web/cgi-irc/
17:29:41 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
17:31:16Nico_Pdamn, why do problems never show up when I need them to? :(
17:32:22alienbiker99i wish i could help you haha.
17:35:12DerPapstbug always vanish when someone wants to kill'em.
17:35:36 Join billenium [0] (
17:36:37 Quit BigBambi ("Leaving")
17:40:55 Join JETC- [0] (
17:41:19 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
17:45:28Nico_Palienbiker99: shouldn't have the bitmap corruptions anymore
17:46:22Nico_PI also encountered another unrelated bug in the process of fixing that one :)
17:46:39Nico_Pbut I have to run, bbs
17:48:49dionoea_Nico_P: is album art support going to make it in svn ?
17:49:03Nico_Pdionoea_: yes
17:49:19Nico_PI've made it work with MoB
17:49:30dionoea_cool :) (although i probably won't have any album art to display anyway)
17:50:05*dionoea_ wonders how long we'll have to wait for a "Please implement coverflow" feature request
17:50:48 Nick dionoea_ is now known as dionoea (
17:51:17Nico_Pdionoea: no reason why it shouldn't be doable on powerful targets ;)
17:51:25billeniumShower BBl
17:52:02alienbiker99Nico_P im building it now for my h300
17:52:14 Join midgey [0] (
17:54:03 Quit J3TC- (".UPP.")
17:54:49Nico_Palienbiker99: cool, thanks. I'm really leaving now, but I'll read the log when I come back (in not too long)
17:55:14alienbiker99crap i got an error
17:55:59 Quit BigBambi ("Leaving")
17:56:45DerPapstdoes this AA path read the AA off the metadata of the file or a seperate bmp?
18:02:00dionoea(or both?)
18:07:20n1sDerPapst: separate bmp at least for now
18:08:19 Join smokinggnu [0] (i=3e2fb665@gateway/web/cgi-irc/
18:09:10DerPapstah ok
18:09:59 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
18:14:58smokinggnuAre you familiar with the vmware debian workspace?
18:15:11BigBambiI have used it in the past
18:15:31smokinggnuis it possible to create shortcuts on the workspace?
18:15:45BigBambiIt uses fluxbox doesn't it?
18:16:28BigBambiI don' know much about fluxbox, you are probably better checking out a fluxbox site, but to my knowledge, no
18:22:15alienbiker99i got it to build. my tree was bad.
18:24:14nanoksmokinggnu: i seem to remeber there was a component of fluxbox which can do that for you
18:24:41nanoknot that i would ever use it. i like my desktop clean ;)
18:28:24alienbiker99erggi need to find a wps for an H300 that just has album art.
18:28:28 Join Jason [0] (i=acca90e2@gateway/web/cgi-irc/
18:28:30smokinggnuit's annoying that I have to
18:28:47smokinggnuenter the same command all the time
18:29:03smokinggnulike "./tools/configure" etc..
18:29:50alienbiker99Nico_P, its hard to find a wps to test that doesnt use scrolling margins
18:29:50 Quit Jason (Client Quit)
18:29:50 Quit Thundercloud (Read error: 104 (Connection reset by peer))
18:29:55n1ssmokinggnu: you should only need to run configure in the same dir once
18:30:08smokinggnuIf I had a shortcut on the desktop, I'd only have to click it and avoid the typing ... yes, I'm lacy. ;)
18:30:38n1s../tools/configure is quicker than grabbing the mouse for me :)
18:30:42smokinggnubut "make", "make install", "./rockboxui"
18:31:05smokinggnuI have to enter that combination quite often!
18:31:09n1smake && make install && ./rockboxui :)
18:31:23n1sput it in a script
18:31:37 Join Arathis [0] (
18:31:39smokinggnuI only know how to make .bat scrips ;)
18:31:59tuplanollajust type it in a text file
18:32:12tuplanollathe give permission to run the script
18:32:20tuplanollawith chmod +x script
18:32:24n1sand then execute the script
18:32:58tuplanollayeah, that is required sometimes :)
18:33:50 Join sarixe [0] (
18:35:59 Join theclaw_ [0] (
18:36:29Nico_Palienbiker99: I just added the tags to icatcher
18:36:45Nico_Pthe AA covers the texts, but for testing I don't care
18:38:11alienbiker99ok. can you upload your wps to pastebin>
18:38:36Nico_Palienbiker99: it's the gigabeat version. but just add the tags like in
18:38:42*Soap suggests a "buy all the devs a voltohmeter fund"
18:38:44alienbiker99ok thanks
18:41:09 Quit maddler ("Lost terminal")
18:41:31 Join maddler [0] (
18:42:45 Quit theclaw (Nick collision from services.)
18:42:48 Nick theclaw_ is now known as theclaw (n=theclaw@unaffiliated/theclaw)
18:43:10alienbiker99errr i must be doing something wrong
18:43:28Nico_Palienbiker99: you need to add two tags
18:44:01Nico_Pl|50|70||| (or whatever) alone on a line, and somewhere
18:44:07 Quit BigBambi ("Leaving")
18:44:25smokinggnuthe script works, but can I save it on the desktop?
18:44:31Nico_P"l|50|70|||" (or whatever) alone on a line, and "" somewhere
18:44:36 Join PaulPosition [0] (
18:45:30Nico_Palienbiker99: I can't manage to write the tags here :p but they are on
18:46:23*Nico_P bbs
18:47:01 Join Buschel [0] (
18:47:56alienbiker99oh wow im stupid. im editing the rwps. grr
18:49:52PaulPositionlol.. viva la remote.
18:51:41 Part smokinggnu
18:53:48 Quit sarixe ("(EE) Failed to load "quit" module")
18:56:42 Join Frazz [0] (
19:02:06 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
19:05:25***Saving seen data "./dancer.seen"
19:05:36 Quit billenium (Read error: 110 (Connection timed out))
19:06:04 Join bertrik [0] (
19:07:12SoapNico_P: I have an idea for AA which works around the scale-an-image-at-runtime issue.
19:07:38Nico_PSoap: yeah?
19:08:02SoapWhat if the AA code would look for different bitmaps based upon the image size as called for in the WPS?
19:08:42Nico_PSoap: this was suggested, but IMHO it's not practical
19:08:42Soapie - if the wps wants a 100x100 image the code first looks for cover100x100.bmp, and if it doesn't find it then goes for cover.bmp.
19:08:56Soapdang - I missed that suggestion - didn't mean to waste time.
19:09:19alienbiker99Nico_P i got it to work. the patch is good
19:09:32Nico_PSoap: no problem ;) thanks for suggesting
19:09:36Nico_Palienbiker99: cool
19:11:40Nico_Pmy testing has been good too
19:12:55 Join jpt9 [0] (
19:13:11jpt9the stereo width and balance controls don't work in the simulator.
19:13:35jpt9neither does the channel configuration menu −− it's always on stereo, regardless of the setting.
19:13:39 Join ilgufo [0] (
19:13:52rasherpreglow: is speex getting unreasonably fast by now?
19:17:00bertrikjpt9: a lot of stuff works differently in the simulator than it does on the real targets, as far as I've seen
19:17:04 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
19:17:13*jpt9 wants to play with it...
19:17:39*jpt9 's sansa should be arriving in about a week.
19:17:41bertriktoo bad really, I wouldn't mind getting rid of some #ifdef SIMULATORs
19:19:15rasherNico_P: Why are you opposed to the multiple bitmaps idea? It would make it possible to switch between themes with different size cover-art without having to re-create the cover images
19:19:27rasherAt least, iiuc
19:24:49Nico_Prasher: it would make the code much more complex for little benefit (IMHO). Instead of testing for the existence of certain known files, we'd have to scan the dir of the track to see which sizes are available and then choose
19:25:29rasherOh? Wouldn't you just see if cover%dx%d.bmp is available before using cover.bmp?
19:25:31Nico_Palso the resizing could be done at loading time, so only once per file
19:25:48Nico_Prasher: what if the WPS uses 105x105?
19:26:02 Join barrywardell [0] (
19:26:11Nico_Pwe'd want to use cover100x100.bmp but how would we know without scanning the dir?
19:26:20bertrikis it that hard to resize?
19:26:25Soapthat would be a problem for the user to solve.
19:26:26Nico_Pbertrik: no
19:26:27rasherNico_P: Ah, I was thinking in the case where no resizing is done
19:26:35Soap(the 105x105) issue.
19:26:53rasherNico_P: then I'd want to look first for cover100x100.bmp (and ONLY that size), then for cover.bmp
19:27:04SoapIt would be up to the user to have an appropriate file named correctly, or suffer the consequences of using the "default" cover.bmp
19:27:13Nico_Phmm yeah then why not
19:27:32rasherTo allow the user to have both cover100x100 and cover50x50, if he switches between two themes that use those sizes
19:27:40Soapbut if resizing is comming - then there is no need.
19:27:55Nico_PI prefer resizing personally
19:28:18SoapOh I do as well, but I'm not offering to do it - so I sure as fuck can't be demanding it be done.
19:28:34rasherDepending how demanding/pretty it is, this method could still be preferable in some situations.
19:28:42 Quit DerPapst ("So Long And Thanks For All The Fish!")
19:29:44Soapwould allow the user to use a different image for very-small "album art" - which might be something totally non CD cover related.
19:30:10Soapwould there be a WPS tag for next-track-AA?
19:30:35rasherBesides, I imagine looking for cover%dx%d.bmp before cover.bmp would be a few lines of code, no?
19:30:38Soapdo a little thumbnail 20x20 cover.bmp of the next track's cover in a corner of a large screen?
19:31:02rasherIn the case where you look strictly for the correct size, that is.
19:32:12SoapI assume an "equal or less than" search would be a PITA.
19:32:12Nico_PSoap: I wasn't imagining next track AA but it would be easy
19:32:45Robin0800Nico_P,can't get the file to patch it has errors
19:32:46 Quit lee-qid (Read error: 110 (Connection timed out))
19:33:00Soapcombined with loading different sized covers, next track AA would be neat.
19:33:12Nico_PRobin0800: can't patch or can't compile?
19:33:45Robin0800Nico_P, can't patch
19:33:50Nico_PSoap: I don't think it's desirable to load more than one bitmap per track though... perhaps not as easy as I thought
19:34:31Soapoh well
19:35:13Nico_PSoap: it all depends on how expensive is the bitmap resizing. I doubt very much that doing it on each display would be good
19:35:17Robin0800Nico_P,not suire how to make patch copied and pasted the saved as was this correct?
19:35:34Nico_PRobin0800: you ned to save the raw text version
19:36:10 Quit BigBambi ("Leaving")
19:36:13Robin0800Nico_P,don't understand
19:36:59Nico_PRobin0800: to copy the patch, you're using text the box at the bottom of the page, right?
19:37:27rasherNico_P: Even with in-rockbox resizing, I'd imagine a pc-side tool would do a better job at resizing, so looking for strict size-matching coverfiles would sitll be a very nice feature imho
19:38:21Nico_Prasher: I think you might be convincing me :p
19:38:53rasherNico_P: Again, I'm not suggesting searching for near-matches. Only checking if the exact match is available - should be fairly cheap to do
19:39:07Nico_Pyeah that should be ok
19:39:49rasherAnd allow some flexibility while resizing is not available, and (probably) slightly better results even when it is
19:41:14barrywardellthe iAudio m5 doesn't have a fm tuner, right?
19:41:19 Quit Echelon (Remote closed the connection)
19:43:45 Join DerPapst [0] (
19:45:00 Quit JRoT|Stage ()
19:45:54 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
19:46:14 Join criznach [0] (
19:50:00Robin0800Nico_P,copied off pastebin pasted it to a text fille saved it as albumart.patch its still failing
19:50:19Nico_PRobin0800: what's the error message?
19:50:35PaulPositionRobin - Pasted to a text file through what? notepad-like app or some text processor?
19:50:53PaulPositionbe wary of wordpad and other FORMATED text editors..
19:55:26rasherNico_P: in playback.c: Hunk #6 succeeded at 3316
19:55:45rasherNot an error, I know - just so you're aware of it
19:55:54rasherWhatever it is.
19:56:01Nico_Prasher: yes?
19:56:20rasherOther than that - it applied like it should for me on a fresh svn checkout
19:56:30Nico_PRobin0800: I'll send another patch
19:56:39rasherNico_P: eh, Hunk #6 succeeded at 3316 (offset 1 line)
19:58:16rasherNico_P: Ah, that would be your latest commit causing that. Missed that
19:58:53 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
19:58:57rasherRobin0800: do you have a fresh and unmodified svn checkout?
19:59:21rasherOr are you using a tarball perhaps?
19:59:26BigBambin1s: cheers
20:00:00Nico_PRobin0800: if it's not my fault
20:00:12Nico_Pif it doesn't work I mean
20:08:04merbananpreglow: patch pending on the mailinglist
20:08:50merbananpreglow: should fix most of the problems
20:09:03merbananpreglow: but not all
20:11:20markunmerbanan: on ffmpeg-devel?
20:12:32merbananSoC mailinglist
20:12:33 Quit BigBambi ("Please insert girder")
20:12:52markunah, of course
20:18:16 Quit barrywardell ()
20:23:48 Join barrywardell [0] (
20:25:08 Quit barrywardell (Client Quit)
20:30:19rasherOh dear, the "hide bar on blank lines" change makes using text_editor on files with blank lines quite challenging
20:37:50 Quit TMM (Remote closed the connection)
20:42:26 Quit Soap (Read error: 104 (Connection reset by peer))
20:42:26 Quit tchan (SendQ exceeded)
20:43:13 Quit maddler (
20:43:13 Quit mrkiko (
20:43:13 Quit miepchen^schlaf (
20:43:13 Quit sslashes (
20:43:13 Quit courtc (
20:43:13 Quit marcosource (
20:43:13 Quit animeloe (
20:43:13 Quit gtkspert (
20:43:13 Quit [mbm] (
20:43:13 Quit BHSPitMonkey (
20:43:13 Quit desowin (
20:43:13 Quit markun (
20:43:13 Quit spiorf (
20:43:13 Quit _jhMikeS_ (
20:43:13 Quit ATravelingGeek (
20:43:13 Quit Dark_Apostrophe (
20:43:13 Quit sd__ (
20:43:13 Quit Frazz (
20:43:13 Quit PaulPosition (
20:43:13 Quit theclaw (
20:43:13 Quit FOAD (
20:43:13 Quit jake_ (
20:43:13 Quit n1s (
20:43:13 Quit ender` (
20:43:13 Quit solatis (
20:43:13 Quit jmspeex (
20:43:13 Quit Arathis (
20:43:13 Quit GnosticE (
20:43:13 Quit merbanan (
20:43:13 Quit HellDragon (
20:43:13 Quit Seedy (
20:43:13 Quit crashd (
20:43:13 Quit merbzt (
20:43:13 Quit Xerion (
20:43:13 Quit DogBoy (
20:43:13 Quit ilgufo (
20:43:13 Quit midgey (
20:43:13 Quit random_desu_is_s (
20:43:13 Quit Toxicity999 (
20:43:13 Quit Bagder (
20:43:13 Quit linuxstb (
20:43:13 Quit SirFunk (
20:43:13 Quit Isolinear (
20:43:13 Quit zicho (
20:43:13 Quit [omni] (
20:43:13 Quit scorche (
20:43:13 Quit kclaf (
20:43:13 Quit jhulst (
20:43:13 Quit midkay (
20:43:13 Quit ivan` (
20:43:13 Quit J (
20:43:13 Quit shodanX (
20:43:13 Quit jurrie (
20:43:13 Quit annulus_ (
20:43:13 Quit TTThomas (
20:43:13 Quit micols (
20:43:13 Quit joshin (
20:43:16 Quit maxkelley (
20:43:16 Quit tedrock (
20:43:16 Quit blithe (
20:43:16 Quit UncleRemus (
20:43:16 Quit RaRe (
20:43:16 Quit jepler (
20:43:16 Quit GodEater (
20:43:16 Quit tuplanolla (
20:43:16 Quit DiDjCodt (
20:43:16 Quit hannesd (
20:43:16 Quit preglow (
20:43:16 Quit lodesi (
20:43:16 Quit fxb__ (
20:43:16 Quit toffe82 (
20:43:16 Quit daurn|laptop (
20:43:16 Quit Ave (
20:43:16 Quit newbyx86 (
20:43:16 Quit jumijoze (
20:43:16 Quit Kohlrabi (
20:43:16 Quit pabs (
20:43:16 Quit feisar (
20:43:16 Quit DerPapst (
20:43:16 Quit jpt9 (
20:43:16 Quit bertrik (
20:43:16 Quit PaulJam (
20:43:16 Quit amiconn (
20:43:16 Quit davina (
20:43:16 Quit iamben_ (
20:43:17 Quit alienbiker99 (
20:43:17 Quit Shaid (
20:43:17 Quit Rick (
20:43:17 Quit andrewg867 (
20:43:17 Quit rvvs89 (
20:43:17 Quit in-jane (
20:43:17 Quit lostlogic (
20:43:17 Quit Lynx_ (
20:43:17 Quit dionoea (
20:43:17 Quit n17ikh (
20:43:17 Quit qwm (
20:43:17 Quit ze (
20:43:17 Quit criznach (
20:43:17 Quit ompaul (
20:43:17 Quit Buschel (
20:43:17 Quit JETC- (
20:43:17 Quit Robin0800 (
20:43:17 Quit moos (
20:43:17 Quit Rob222241 (
20:43:17 Quit thegeek_ (
20:43:17 Quit advcomp2019 (
20:43:17 Quit nanok (
20:43:17 Quit gromit` (
20:43:17 Quit DataGhost (
20:43:17 Quit sbeh (
20:43:17 Quit lastebil (
20:43:17 Quit XavierGr (
20:43:17 Quit crwll (
20:43:17 Quit parafin (
20:43:17 Quit krazykit (
20:43:17 Quit sneakums (
20:43:17 Quit chris__ (
20:43:17 Quit Nimdae (
20:43:17 Quit BjoernErik (
20:43:17 Quit bnakiddmj (
20:43:17 Quit tierra (
20:43:17 Quit Hadaka (
20:43:17 Quit Jon-Kha (
20:43:17 Quit rasher (
20:43:17 Quit Slasheri (
20:43:30 Join animeloe_ [0] (
20:43:38 Join tchan [0] (
20:43:50NJoinhannesd [0] (
20:43:50 Join Blue_Alien [0] (
20:43:50 Join Soap_ [0] (
20:43:50NJoincriznach [0] (
20:43:50NJoinompaul [0] (n=ompaul@gnewsense/friend/ompaul)
20:43:50NJoinDerPapst [0] (
20:43:50NJoinilgufo [0] (
20:43:50NJoinjpt9 [0] (
20:43:50NJoinbertrik [0] (
20:43:50NJoinFrazz [0] (
20:43:50NJoinBuschel [0] (
20:43:50NJoinPaulPosition [0] (
20:43:50NJoinmaddler [0] (
20:43:50 Join theclaw [0] (n=theclaw@unaffiliated/theclaw)
20:43:50NJoinArathis [0] (
20:43:50NJoinmidgey [0] (
20:43:50NJoinJETC- [0] (
20:43:50NJointoffe82 [0] (
20:43:50NJoinrandom_desu_is_s [0] (
20:43:50NJoinFOAD [0] (
20:43:50NJoinPaulJam [0] (
20:43:50NJoinToxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
20:43:50NJoinRobin0800 [0] (
20:43:50 Join desowin [0] (n=desowin@unaffiliated/desowin)
20:43:50NJoinmrkiko [0] (
20:43:50NJoinmarkun [0] (n=markun@rockbox/developer/markun)
20:43:50NJoinGnosticE [0] (
20:43:50NJoinamiconn [0] (n=jens@rockbox/developer/amiconn)
20:43:50NJoinUncleRemus [0] (
20:43:50NJoinmoos [0] (
20:43:50NJoinspiorf [0] (
20:43:50NJoin_jhMikeS_ [0] (n=jethead7@rockbox/developer/jhMikeS)
20:43:50NJoinjake_ [0] (
20:43:50NJointierra [0] (
20:43:50 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
20:43:50NJoinn1s [0] (
20:43:50NJoindavina [0] (
20:43:50NJoinmerbanan [0] (n=banan@
20:43:50NJoinATravelingGeek [0] (n=ATG@pdpc/supporter/student/ATravelingGeek)
20:43:50NJoindaurn|laptop [0] (n=fake@unaffiliated/daurnimator)
20:43:50NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
20:43:50NJoinRob222241 [0] (
20:43:50NJoinender` [0] (
20:43:50NJoinSirFunk [0] (
20:43:50NJoinmiepchen^schlaf [0] (
20:43:50 Join Dark_Apostrophe [0] (n=darkapos@supporter/monthlybyte/DarkApostrophe)
20:43:50NJointhegeek_ [0] (
20:43:50NJoinHellDragon [0] (i=jd@unaffiliated/helldragon)
20:43:50NJoiniamben_ [0] (
20:43:50NJoinIsolinear [0] (
20:43:50NJoinSeedy [0] (
20:43:50NJoinalienbiker99 [0] (
20:43:50NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
20:43:50NJoinnanok [0] (n=nanok@
20:43:50NJoingromit` [0] (
20:43:50NJoinsd__ [0] (n=sd@
20:43:50NJoinmerbzt [0] (
20:43:50NJoinzicho [0] (
20:43:50NJoinDataGhost [0] (
20:43:50NJoinShaid [0] (
20:43:50NJoin[omni] [0] (
20:43:50NJoincrashd [0] (
20:43:50NJoinscorche [0] (i=Blah@rockbox/administrator/scorche)
20:43:50NJoinkclaf [0] (n=kclaf@
20:43:50NJoinjhulst [0] (n=jhulst@unaffiliated/jhulst)
20:43:50NJoinmidkay [0] (n=midkay@rockbox/developer/midkay)
20:43:50 Join Rick [0] (i=rick@unaffiliated/rick)
20:43:50NJoincourtc [0] (n=court@unaffiliated/courtc)
20:43:50 Join [mbm] [0] (i=mbm@openwrt/developer/mbm)
20:43:50NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
20:43:50NJoingtkspert [0] (
20:43:50NJoinsslashes [0] (i=sslashes@
20:43:50NJoinmarcosource [0] (
20:43:50NJoinanimeloe [0] (n=animeloe@unaffiliated/animeloe)
20:43:50NJoinsolatis [0] (
20:43:50NJoinjmspeex [0] (
20:43:50 Join ivan` [0] (n=ivan`@unaffiliated/ivan/x-000001)
20:43:50NJoinNimdae [0] (
20:43:50NJoinjepler [0] (n=jepler@emc/developer/jepler)
20:43:50NJoinannulus_ [0] (
20:43:50NJoinDogBoy [0] (n=john@unaffiliated/dogboy)
20:43:50NJoinXerion [0] (
20:43:50NJoinLynx_ [0] (
20:43:50NJoinTTThomas [0] (
20:43:50NJoinqwm [0] (
20:43:50NJoinchris__ [0] (
20:43:50NJointedrock [0] (
20:43:50NJoinsneakums [0] (
20:43:50NJoinze [0] (
20:43:50NJoinfeisar [0] (
20:43:50NJoinkrazykit [0] (
20:43:50NJoinparafin [0] (
20:43:50NJoincrwll [0] (
20:43:50NJoinbnakiddmj [0] (i=0@
20:43:50NJoinXavierGr [0] (
20:43:50NJoinBjoernErik [0] (
20:43:50NJoinlastebil [0] (
20:43:50NJoinsbeh [0] (n=sbeh@
20:43:50NJoinlostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
20:43:50NJoinKohlrabi [0] (
20:43:50NJoinpabs [0] (
20:43:50NJoinin-jane [0] (i=jane@
20:43:50NJoinmaxkelley [0] (
20:43:50NJoindionoea [0] (
20:43:50NJoinjumijoze [0] (n=jumijoze@unaffiliated/mike-zed)
20:43:50NJoinrvvs89 [0] (n=rvvs89@pdpc/supporter/active/rvvs89)
20:43:50NJoinjoshin [0] (n=joshin@unaffiliated/joshin)
20:43:50NJoinmicols [0] (
20:43:50NJoinDiDjCodt [0] (
20:43:50NJoinjurrie [0] (
20:43:50NJoinblithe [0] (
20:43:50NJoinshodanX [0] (
20:43:50NJointuplanolla [0] (
20:43:50NJoinJ [0] (
20:43:50NJoinRaRe [0] (
20:43:50NJoinGodEater [0] (n=bryan@rockbox/staff/GodEater)
20:43:50NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
20:43:50NJoinrasher [0] (n=rasher@rockbox/developer/rasher)
20:43:50NJoinJon-Kha [0] (
20:43:50NJoinHadaka [0] (
20:43:50NJoinfxb__ [0] (
20:43:50NJoinlodesi [0] (
20:43:50NJoinpreglow [0] (n=thomj@rockbox/developer/preglow)
20:43:50NJoinAve [0] (
20:43:50NJoinnewbyx86 [0] (
20:43:50NJoinn17ikh [0] (
20:43:50NJoinandrewg867 [0] (
20:43:51***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! ( )'
20:43:57***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! ( )'
20:44:08 Quit animeloe (
20:44:08 Quit marcosource (
20:44:08 Quit gtkspert (
20:44:08 Quit sslashes (
20:44:08 Quit mrkiko (
20:44:08 Quit miepchen^schlaf (
20:44:08 Quit courtc (
20:44:08 Quit BHSPitMonkey (
20:44:08 Quit [mbm] (
20:44:08 Quit maddler (
20:44:10Blue_Alieni just installed rockbox on my 60 gig ipod and i was wondering if there is a way i can watch my videos onit?
20:44:12NJoinmaddler [0] (
20:44:12NJoinmrkiko [0] (
20:44:12NJoinmiepchen^schlaf [0] (
20:44:12NJoinmarcosource [0] (
20:44:12NJoinsslashes [0] (i=sslashes@
20:44:12NJoingtkspert [0] (
20:44:12NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
20:44:12NJoin[mbm] [0] (i=mbm@openwrt/developer/mbm)
20:44:12NJoincourtc [0] (n=court@unaffiliated/courtc)
20:44:12 Join theclaw_ [0] (
20:44:12 Nick Soap_ is now known as Soap (
20:44:14Blue_Alien just installed rockbox on my 60 gig ipod and i was wondering if there is a way i can watch my videos onit?
20:44:22 Quit animeloe_ (Client Quit)
20:44:27 Quit PaulPosition ()
20:44:36 Join animeloe_ [0] (
20:44:53DerPapstwt.. what was that message?
20:45:04 Quit animeloe_ (Client Quit)
20:45:16 Join animeloe_ [0] (
20:45:23Blue_Alien just installed rockbox on my 60 gig ipod and i was wondering if there is a way i can watch my videos onit?
20:45:30SoapBlue_Alien: see the manual and the wiki on PluginMpegplayer
20:45:52Soapand in the future - please be more paitent than asking the same question three times in 68 seconds.
20:46:00 Quit animeloe_ (Client Quit)
20:46:00ompaulDerPapst, referred to as a netsplit, nothing to worry about, leave it to network admins to get their kit together
20:46:11 Join animeloe_ [0] (n=animeloe@unaffiliated/animeloe)
20:46:21DerPapstompaul: ah ok :)
20:46:50 Quit animeloe_ (Client Quit)
20:47:13CtcpVersion from freenode-connect!freenode@freenode/bot/connect
20:47:20 Quit Blue_Alien (Client Quit)
20:47:20NJoinanimeloe [0] (n=animeloe@unaffiliated/animeloe)
20:48:43 Quit jpt9 ("ChatZilla [Firefox]")
20:50:07 Join Thundercloud [0] (
20:54:17 Join theclaw__ [0] (
20:56:46 Quit jhulst (No route to host)
20:57:03 Quit TTThomas (Read error: 113 (No route to host))
21:00:59 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
21:01:11 Quit theclaw (Read error: 110 (Connection timed out))
21:05:13 Join theclaw [0] (
21:05:26***Saving seen data "./dancer.seen"
21:06:10 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
21:06:24 Quit advcomp2019 (Nick collision from services.)
21:06:28 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
21:07:49 Join hannesd_ [0] (
21:12:05 Join Jeton [0] (n=chatzill@
21:13:59 Join linuxstb_ [0] (n=chatzill@rockbox/developer/linuxstb)
21:14:37 Quit theclaw_ (Read error: 101 (Network is unreachable))
21:14:41 Quit hannesd (Connection timed out)
21:15:49 Join FOAD_ [0] (
21:16:24 Quit Arathis ("Bye, bye")
21:17:25 Quit theclaw__ (Read error: 110 (Connection timed out))
21:17:28 Quit karashata ("Leaving.")
21:18:41Nico_Prasher: I did the size-in-name thing
21:19:52 Quit Frazz ("Leaving")
21:20:04Nico_PI need to make a little bit less hackish now
21:26:16 Join Frazz [0] (
21:27:33linuxstb_What filename does it look for?
21:28:17Soapwoo hoo!
21:28:26 Quit hannesd_ (Connection timed out)
21:31:08 Quit amiconn (Nick collision from services.)
21:31:16 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
21:31:52Soaplinuxstb: if Nico_P isn't going to answer... The discussion previously was on it looking for a file named the same as the size called for in the WPS, and then if that is not found, the default cover.bmp.
21:32:18Soapie - if the WPS calls for 100x100 album art the first file looked for would be 100x100cover.bmp
21:32:19Nico_Plinuxstb_: something like cover.100x100.bmp then cover.bmp
21:32:45 Quit FOAD (Read error: 110 (Connection timed out))
21:32:45 Nick FOAD_ is now known as FOAD (
21:33:05Nico_Plinuxstb_: but there are other possible filenames... trackname.bmp, albumname.bmp, cover.bmp, then the same in the parent dir
21:33:11Nico_Pmight be a little overkill
21:33:30linuxstb_OK, I was just curious exactly what filename you had chosen. cover.WxH.bmp seems perfect.
21:34:01 Join stevesel [0] (
21:34:17 Part stevesel ("Konversation terminated!")
21:37:42linuxstb_Are there any WPSs which use album art but don't need any other patches? I was thinking that once you commit it, it could be nice to have a WPS in SVN that can use it...
21:40:07Nico_Plinuxstb_: on the gigabeat and sansa there probably are some, but on targets with landscape screens, it's hard to make an AA WPS without scrolling margins
21:44:07Nico_Plinuxstb_: but we could adapt one and simply disable the scrolling, although that's not very elegant
21:45:10Nico_Pthe WPS should also use the conditional tag allowing for different layouts when there is AA and when there is not
21:45:58nanokNico_P: uhm,the sansa i think is portrait (don;t know about the gigabeat though)
21:46:23krazykitnanok, yes, that's what he was implying
21:47:06Nico_Pyeah and that's why it's easier to make an AA WPS without scrolling margins... you rarely want to put text next to the AA on those targets
21:47:09 Join sdubois92 [0] (
21:47:19sdubois92does songbird support rockbox?
21:47:40rashersdubois92: I suggest you ask a songbird forum. What do you mean by support though?
21:48:22sdubois92like how itunes automatically puts music on the ipod. does songbird do the same for a rockbox device
21:48:37nanokwell, anyway, to be honest idon;t care much for album art. i discarded the idea straight on, especially that those wps's need(ed) custombuilds. i am more tempted by other (mostly nonvisual) features
21:48:52nanokso i don't know much about albumart
21:49:11nanokbut, ofcourse, that's just me. overall i think it's great that now it will be merged
21:49:50linuxstb_sdubois92: A quick google points here -
21:49:52rashersdubois92: Again, this is not the forum for such a question. Rockbox devices are simply a regular USB drive.
21:50:35 Quit tedrock (Client Quit)
21:52:29 Quit criznach ("ChatZilla [Firefox]")
21:52:45 Quit ilgufo ("So Long, and Thanks For All the Fish -")
21:53:27UncleRemus I tested my iPod (Photo, 4th G) with rockbox today and noticed that the battery only played for 1:15 hours. I booted the Apple firmware and it still plays after 3 hours and the battery seems like 70-80%. Are there any major stuff i could do to improve battery time with rockbox since I reeaally want to use it?
21:55:31linuxstb_I tested my ipod photo last week, and got about 6h 50m playing back MP3s.
21:55:32JetonNico_P: will the resizing of the Album Art .bmp file cause it to be jaggy? Or the resized image will be "smooth"?
21:56:07 Join TPM_Debian [0] (
21:56:20Nico_PJeton: there'll be no resizing in the initial commit, but once I implement it I plan on using a better resizing algorithm than the patch
21:56:32TPM_Debianhello @all
21:57:23JetonNico_P: that's great, since the actual algorithm does make the images appear jagged (you probably knew that ) .
21:57:26Nico_PJeton: but there is already support for files like cover.100x100.bmp or cover.75x75.bmp
21:57:41UncleRemuslinuxstb_: I waould be very happy if I could reach that, how old is your iPod?
21:58:03Jetonyeah, but every theme that supports AA has different sizes specified ,that's why it's hard to keep a single size.
21:58:04Nico_PJeton: yeah, it's a pretty bad algorithm AFAIK
21:58:24Nico_PJeton: what I said above means that you can have several size
21:58:37linuxstb_UncleRemus: I bought it a few months before it was replaced by the 5g - August 2005 I think.
21:59:06UncleRemusOK, seems the same as mine approx
21:59:22Jetoni see, but we wouldn't want 3 AA images(or more) for a song, or album. :)
21:59:28 Quit Frazz ("Leaving")
21:59:31TPM_Debiani got a problem whis my sansa .... the bootloader and firmware are totaly corrupt ... so .. i use e200tool, but this tool prints no device found ... anybody a idea? (sry. my bad english)
21:59:42Nico_PJeton: then wait for resizing ;)
21:59:56UncleRemuslinuxstb_: I was out walking with it when I tested it, can that account for the big difference?
22:00:19Jetonyes sir! :D , and hopefully jpeg (or even better embedded) support ;)
22:01:00Nico_PJeton: those are the steps after resizing
22:01:28linuxstb_UncleRemus: Why do you think that would matter?
22:01:36Nico_PI'm not sure I'm very interested in them though
22:01:53Nico_PUncleRemus: depends on how you walk :p
22:02:34JetonNico_P: well, all good then, The progress made so far is impressive. (thinking how my Sansa was before installing rockbox)
22:02:48UncleRemuslinuxstb_: Maybe I'm just clutching at straws. So, you think your battery is approx. the same as with Apple?
22:03:12TPM_Debiannoone a idea???
22:03:34linuxstb_I never use the Apple firmware, but no, I would expect it to be much better in Apple's firmware - there are known issues with Rockbox drawing more power than it should on some ipods, including the Photo.
22:04:52linuxstb_TPM_Debian: I assume you're running Debian? Are you running e200tool as root? Is your e200 in manufacturing mode?
22:05:18TPM_Debian1.) + 2.) yes ... 3.) i hope it ...
22:06:25TPM_Debianif i use "init" command, so it will print
22:06:29TPM_DebianSearching for device 0781:0720 ... found!
22:06:29TPM_DebianFailed to claim the interface (-1, Operation not permitted)
22:06:29TPM_DebianSearching for device 0b70:0003 ... 9 8 7 6 5 4 3 2 1 0 not found!
22:07:08linuxstb_That "Operation not permitted" looks to be the problem - are you sure you're root?
22:07:13TPM_Debianso .. i dont know, ... found it or not?!? ... well ... i use i2cprogram only the last line will show
22:07:50TPM_Debiana .. lol ...
22:07:59TPM_Debiani was kickt out of su
22:09:05TPM_Debianit will works now .. thx so much :]
22:09:54 Quit Jeton ("ChatZilla [Firefox]")
22:11:15TPM_Debianwell ... e200tool has write the data, but nothin will do on my sansa ... black display on normal / recorery and manufacturing mode
22:12:16 Join Askyron [0] (
22:12:33krazykitAskyron, don't spam and get the hell out.
22:13:04 Part Askyron
22:13:08 Quit GnosticE ()
22:13:18rasherSurprisingly polite, all things considered.
22:13:29krazykiti'm more surprised that it actually worked.
22:14:51 Join billenium [0] (
22:14:59 Quit sdubois92 ("Changing server")
22:16:28 Join hannesd [0] (
22:20:48 Join Arathis [0] (
22:22:19 Quit Toxicity999 (Remote closed the connection)
22:23:35SoapUncleRemus: The best guess is that Rockbox should have around 60-65% of the battery life of Apple firmware on the Apple Photo
22:25:04TPM_Debianhmm ... after init, i2c and recover will short show the sandisk-logo then print it "hold-key locked ... system shutdown" ... but .. if i will start sansa normal or in recovery the display will be blank ....
22:25:14UncleRemusOK, Soap. Is the battery life issues of high importance for the devs? What could be the cause of the discrepancy? Is it not optimal usage of hardware acceleration?
22:25:57 Join Toxicity999 [0] (
22:27:59 Quit Mouser_X ("CGI:IRC (Session timeout)")
22:29:58SoapThe cause is unknown. The iPods use an undocumented chip. The problem is most likely NOT inefficient code - as outside the PP5020 series of chips (later rockbox supported iPods and Sansas) rockbox acheives better than stock battery life using the same code as on the targets it falls short of.
22:30:28SoapThe issue is of high importance - but the options are limited and difficult for such a situation.
22:30:33rasherUncleRemus: probably hardware that's not initialized correctly (or should be put to sleep)
22:31:45jmspeexfreqmod_nx: does your file play in standard speexdec?
22:36:36linuxstb_Bagder: Around?
22:36:40 Quit hannesd (
22:36:40 Quit mrkiko (
22:36:40 Quit sslashes (
22:36:40 Quit courtc (
22:36:40 Quit marcosource (
22:36:40 Quit gtkspert (
22:36:40 Quit BHSPitMonkey (
22:36:40 Quit miepchen^schlaf (
22:36:40 Quit [mbm] (
22:36:40 Quit maddler (
22:36:40 Quit amiconn (
22:36:40 Quit advcomp2019 (
22:36:40 Quit animeloe (
22:36:40 Quit Dark_Apostrophe (
22:36:40 Quit _jhMikeS_ (
22:36:40 Quit spiorf (
22:36:40 Quit markun (
22:36:40 Quit sd__ (
22:36:40 Quit ATravelingGeek (
22:36:40 Quit desowin (
22:36:40 Quit n1s (
22:36:40 Quit jake_ (
22:36:40 Quit ender` (
22:36:40 Quit solatis (
22:36:40 Quit jmspeex (
22:36:40 Quit crashd (
22:36:40 Quit Seedy (
22:36:40 Quit merbzt (
22:36:40 Quit merbanan (
22:36:40 Quit Xerion (
22:36:40 Quit DogBoy (
22:36:40 Quit HellDragon (
22:36:40 Quit Toxicity999 (
22:36:40 Quit FOAD (
22:36:40 Quit jhulst (
22:36:40 Quit J (
22:36:40 Quit tuplanolla (
22:36:40 Quit shodanX (
22:36:40 Quit jurrie (
22:36:40 Quit annulus_ (
22:36:40 Quit kclaf (
22:36:40 Quit Isolinear (
22:36:40 Quit random_desu_is_s (
22:36:40 Quit midgey (
22:36:40 Quit linuxstb (
22:36:40 Quit micols (
22:36:40 Quit joshin (
22:36:40 Quit zicho (
22:36:40 Quit maxkelley (
22:36:40 Quit blithe (
22:36:40 Quit UncleRemus (
22:36:40 Quit ivan` (
22:36:40 Quit RaRe (
22:36:40 Quit jepler (
22:36:40 Quit Bagder (
22:36:40 Quit scorche (
22:36:40 Quit midkay (
22:36:42 Quit DiDjCodt (
22:36:42 Quit [omni] (
22:36:42 Quit SirFunk (
22:36:42 Quit GodEater (
22:36:42 Quit preglow (
22:36:42 Quit lodesi (
22:36:42 Quit fxb__ (
22:36:42 Quit daurn|laptop (
22:36:42 Quit toffe82 (
22:36:42 Quit Ave (
22:36:42 Quit newbyx86 (
22:36:42 Quit jumijoze (
22:36:42 Quit Kohlrabi (
22:36:42 Quit pabs (
22:36:42 Quit feisar (
22:36:42 Quit TPM_Debian (
22:36:42 Quit andrewg867 (
22:36:42 Quit rvvs89 (
22:36:42 Quit in-jane (
22:36:42 Quit lostlogic (
22:36:42 Quit Lynx_ (
22:36:42 Quit Shaid (
22:36:42 Quit PaulJam (
22:36:42 Quit bertrik (
22:36:42 Quit DerPapst (
22:36:42 Quit alienbiker99 (
22:36:42 Quit dionoea (
22:36:42 Quit n17ikh (
22:36:42 Quit iamben_ (
22:36:42 Quit davina (
22:36:42 Quit Rick (
22:36:42 Quit qwm (
22:36:42 Quit ze (
22:36:42 Quit Arathis (
22:36:42 Quit theclaw (
22:36:42 Quit sbeh (
22:36:42 Quit lastebil (
22:36:42 Quit XavierGr (
22:36:42 Quit crwll (
22:36:42 Quit parafin (
22:36:42 Quit krazykit (
22:36:42 Quit sneakums (
22:36:42 Quit chris__ (
22:36:42 Quit Nimdae (
22:36:42 Quit nanok (
22:36:43 Quit thegeek_ (
22:36:43 Quit Rob222241 (
22:36:43 Quit Robin0800 (
22:36:43 Quit JETC- (
22:36:43 Quit Buschel (
22:36:43 Quit moos (
22:36:43 Quit Soap (
22:36:43 Quit BjoernErik (
22:36:43 Quit bnakiddmj (
22:36:43 Quit tierra (
22:36:43 Quit ompaul (
22:36:43 Quit DataGhost (
22:36:43 Quit gromit` (
22:36:44 Quit Hadaka (
22:36:44 Quit Jon-Kha (
22:36:44 Quit rasher (
22:36:44 Quit Slasheri (
22:37:48*freqmod_nx is checking if it plays
22:40:10freqmod_nxDecoding 16000 Hz audio using wideband (sub-band CELP) mode (stereo, VBR)
22:40:24freqmod_nxit works but it is /\/\
22:41:18freqmod_nxprobably a stereo error
22:42:15 Quit guyzmo (Remote closed the connection)
22:43:23billeniumis anyone here?
22:43:37 Join przemhb [0] (n=chatzill@
22:43:39 Quit billenium ("Ex-Chat")
22:51:46 Join Llorean [0] (n=llorean@
22:54:43NJoinToxicity999 [0] (
22:54:43NJoinArathis [0] (
22:54:43NJoinhannesd [0] (
22:54:43NJoinTPM_Debian [0] (
22:54:43NJoinamiconn [0] (n=jens@rockbox/developer/amiconn)
22:54:43NJoinFOAD [0] (
22:54:43NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
22:54:43 Join theclaw [0] (n=theclaw@unaffiliated/theclaw)
22:54:43NJoinjhulst [0] (n=jhulst@unaffiliated/jhulst)
22:54:43NJoinanimeloe [0] (n=animeloe@unaffiliated/animeloe)
22:54:43NJoincourtc [0] (n=court@unaffiliated/courtc)
22:54:43NJoin[mbm] [0] (i=mbm@openwrt/developer/mbm)
22:54:43NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
22:54:43NJoingtkspert [0] (
22:54:43NJoinsslashes [0] (i=sslashes@
22:54:43NJoinmarcosource [0] (
22:54:43NJoinmiepchen^schlaf [0] (
22:54:43NJoinmrkiko [0] (
22:54:43NJoinmaddler [0] (
22:54:43 Join Soap [0] (n=Soap@rockbox/staff/soap)
22:54:43NJoinompaul [0] (n=ompaul@gnewsense/friend/ompaul)
22:54:43NJoinDerPapst [0] (
22:54:43NJoinbertrik [0] (
22:54:43NJoinBuschel [0] (
22:54:43NJoinmidgey [0] (
22:54:43NJoinJETC- [0] (
22:54:43NJointoffe82 [0] (
22:54:43NJoinrandom_desu_is_s [0] (
22:54:43NJoinPaulJam [0] (
22:54:43NJoinRobin0800 [0] (
22:54:43NJoindesowin [0] (n=desowin@unaffiliated/desowin)
22:54:43NJoinmarkun [0] (n=markun@rockbox/developer/markun)
22:54:43NJoinUncleRemus [0] (
22:54:43NJoinmoos [0] (
22:54:43NJoinspiorf [0] (
22:54:43NJoin_jhMikeS_ [0] (n=jethead7@rockbox/developer/jhMikeS)
22:54:43NJoinjake_ [0] (
22:54:43NJointierra [0] (
22:54:43NJoinBagder [0] (n=daniel@rockbox/developer/bagder)
22:54:43NJoinn1s [0] (
22:54:43NJoindavina [0] (
22:54:43NJoinmerbanan [0] (n=banan@
22:54:43NJoinATravelingGeek [0] (n=ATG@pdpc/supporter/student/ATravelingGeek)
22:54:43NJoindaurn|laptop [0] (n=fake@unaffiliated/daurnimator)
22:54:43NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
22:54:43NJoinRob222241 [0] (
22:54:43NJoinender` [0] (
22:54:43NJoinSirFunk [0] (
22:54:43NJoinDark_Apostrophe [0] (n=darkapos@supporter/monthlybyte/DarkApostrophe)
22:54:43NJointhegeek_ [0] (
22:54:43NJoinHellDragon [0] (i=jd@unaffiliated/helldragon)
22:54:43NJoiniamben_ [0] (
22:54:43NJoinIsolinear [0] (
22:54:43NJoinSeedy [0] (
22:54:43NJoinalienbiker99 [0] (
22:54:43NJoinnanok [0] (n=nanok@
22:54:43NJoingromit` [0] (
22:54:43NJoinsd__ [0] (n=sd@
22:54:43NJoinmerbzt [0] (
22:54:43NJoinzicho [0] (
22:54:43NJoinDataGhost [0] (
22:54:43NJoinShaid [0] (
22:54:43NJoin[omni] [0] (
22:54:43NJoincrashd [0] (
22:54:43NJoinscorche [0] (i=Blah@rockbox/administrator/scorche)
22:54:43NJoinkclaf [0] (n=kclaf@
22:54:43NJoinmidkay [0] (n=midkay@rockbox/developer/midkay)
22:54:43NJoinRick [0] (i=rick@unaffiliated/rick)
22:54:43NJoinsolatis [0] (
22:54:43NJoinjmspeex [0] (
22:54:43NJoinivan` [0] (n=ivan`@unaffiliated/ivan/x-000001)
22:54:43NJoinNimdae [0] (
22:54:43NJoinjepler [0] (n=jepler@emc/developer/jepler)
22:54:43NJoinannulus_ [0] (
22:54:43NJoinDogBoy [0] (n=john@unaffiliated/dogboy)
22:54:43NJoinXerion [0] (
22:54:43NJoinLynx_ [0] (
22:54:43NJoinqwm [0] (
22:54:43NJoinchris__ [0] (
22:54:43NJoinsneakums [0] (
22:54:43NJoinze [0] (
22:54:43NJoinfeisar [0] (
22:54:43NJoinkrazykit [0] (
22:54:43NJoinparafin [0] (
22:54:43NJoincrwll [0] (
22:54:43NJoinbnakiddmj [0] (i=0@
22:54:43NJoinXavierGr [0] (
22:54:43NJoinBjoernErik [0] (
22:54:43NJoinlastebil [0] (
22:54:43NJoinsbeh [0] (n=sbeh@
22:54:43NJoinlostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
22:54:43NJoinKohlrabi [0] (
22:54:43NJoinpabs [0] (
22:54:43NJoinin-jane [0] (i=jane@
22:54:43NJoinmaxkelley [0] (
22:54:43NJoindionoea [0] (
22:54:43NJoinjumijoze [0] (n=jumijoze@unaffiliated/mike-zed)
22:54:43NJoinrvvs89 [0] (n=rvvs89@pdpc/supporter/active/rvvs89)
22:54:43NJoinjoshin [0] (n=joshin@unaffiliated/joshin)
22:54:43NJoinmicols [0] (
22:54:43NJoinDiDjCodt [0] (
22:54:43NJoinjurrie [0] (
22:54:43NJoinblithe [0] (
22:54:43NJoinshodanX [0] (
22:54:43NJointuplanolla [0] (
22:54:43NJoinJ [0] (
22:54:43NJoinRaRe [0] (
22:54:43NJoinGodEater [0] (n=bryan@rockbox/staff/GodEater)
22:54:43NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
22:54:43NJoinrasher [0] (n=rasher@rockbox/developer/rasher)
22:54:43NJoinJon-Kha [0] (
22:54:43NJoinHadaka [0] (
22:54:43NJoinfxb__ [0] (
22:54:43NJoinlodesi [0] (
22:54:43NJoinpreglow [0] (n=thomj@rockbox/developer/preglow)
22:54:43NJoinAve [0] (
22:54:43NJoinnewbyx86 [0] (
22:54:43NJoinn17ikh [0] (
22:54:43NJoinandrewg867 [0] (
22:54:44***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
22:59:09 Quit ompaul (Client Quit)
23:00:26UncleRemusThanks for your replies, Soap and rasher. Interesting indeed.
23:00:29amiconnNico_P: I think we *will* need a way to easily disable AA. On mono targets and other low-mem targets it doesn't make sense
23:00:30 Join guyzmo [0] (
23:00:31amiconnOn greyscale targets it's to be decided. Might look acceptable with dithering
23:00:32CtcpVersion from freenode-connect!freenode@freenode/bot/connect
23:00:33UncleRemusSo, if I should ever by a new iPod, will the newer ones have better lifetime ratio (i.e. compared to Apple)?
23:00:34amiconnI wonder how you intend to handle targets with more than one LCD...
23:00:34 Join billenium [0] (
23:00:36amiconn...especially as you're planning to go without scaling, and use bmp
23:00:40krazykitUncleRemus, no, they share the issues.
23:01:14LloreanWhy are we going without scaling?
23:01:17UncleRemusHmmm, stupid iPods. :D What devices are known to work really good with rockbox in battery aspects?
23:01:18amiconnEverything coldfire based
23:01:18krazykiteverything except 4+gen ipods and sansas
23:01:19amiconnkrazykit: Not correct
23:01:19krazykitand the h10, right?
23:01:19amiconnEverything that's not portalplayer based
23:01:20LloreanIsn't there a bitmap scaling patch already?
23:01:20DBUGEnqueued KICK amiconn
23:01:20amiconnMore precisely, everything that's not portalplayer PP502x based
23:01:21amiconnThe 5002 is on par with OF
23:01:22 Join GX9900WZ [0] (
23:01:22***Alert Mode level 1
23:01:22amiconnUncleRemus: If you want really long battery runtime, my recommendation would be the iaudio X5L or M5L
23:01:24UncleRemusThanks, I'll go and check it now
23:01:27GX9900WZhi, i just installed rockbox on my sansa e250, i connected it to the computer, but the computer is not recognizing the device
23:01:27amiconnThe M5L is holding the rockbox battery runtime record. Almost 52 hours with lame −−preset standard mp3. Of course a synthetic runtime test is always better than real usage, but still...
23:01:27advcomp2019GX9900WZ, are you in the OF or rockbox
23:01:28bertrikGX9900WZ: USB for sansa e200 does not work yet in rockbox
23:01:29advcomp2019that is why rockbox does not have usb support yet.. so boot into OF
23:01:30GX9900WZoh, so i should switch to the OF when i connect to the computer to upload songs?
23:01:30GX9900WZokay, thanks
23:01:31bertrikto charge the battery and to use USB, you need to use the original firmware
23:01:42bertrikGX9900WZ: convenient way to boot into the original firmware is to turn your sansa off, then plug in the USB cable, it will wake up automatically and go to the OF, annoying database refresh will be suppressed by the rockbox bootloader
23:01:43TPM_DebianoO ... the bat will not charging when the os this does not support?!?! ... wtf?
23:01:44amiconnAfaik the sansa charges from usb
23:01:44bertrikTPM_Debian: it will charge, but very slowly
23:01:45amiconnIn order to draw more than 100mA from an usb port, this must be requested from the host as by usb standards
23:01:45TPM_Debiana ... ok
23:01:45amiconnAnd since rockbox can't do usb yet on pp, we can't request that extra power either
23:01:46bertrikamiconn: in order to draw 100 mA, you need to tell the charger circuit to draw 100 mA, otherwise it defaults to 50 mA AFAIK
23:01:46TPM_Debianmax is a half amp
23:01:47amiconnTPM_Debian: Yes, but only if you tell the host
23:01:47TPM_Debianyes ... i know ;)
23:01:47DBUGEnqueued KICK TPM_Debian
23:01:47TPM_Debianwell ... i must mount my sansa on linux ... i hope i can do that ....
23:01:48amiconnI know there are devices which don't care, but that's violating specs, and some hosts play safe and simply sht down the port in such a case
23:01:48TPM_Debianits my first time
23:01:48GX9900WZthanks for the tip
23:01:48 Quit midkay (Read error: 113 (No route to host))
23:01:49krazykitTPM_Debian, you can mount it in the OF no problem.
23:01:49UncleRemusamiconn: I found the X5L . Will consider it for my next mp3 player.
23:01:50GX9900WZhmm, my sansa is waking up automatically when i plug in the usb
23:01:50krazykitGX9900WZ, yes, that's desired behavior
23:01:50amiconnUncleRemus: That's the plain X5 (60GB model), not the X5L
23:01:50 Quit n17ikh ()
23:01:50GX9900WZi mean, it is not waking up automatically
23:01:50 Quit desowin ("use linux")
23:01:50SoapUncleRemus: don't consider it for too long, they are quickly disapearing.
23:01:50UncleRemusYeah, just noticed it was sold out...
23:01:50 Quit ATravelingGeek (Connection timed out)
23:01:50TPM_Debianwell .. my sansa is corrupt ... and its so very very hard to flash ned system on it .. after i write the i2c.bin ... i need to do (before the player restart) to push the rec button to init the device into recoery mode .. the i need to push the os-image on the device ... befor i can disconnect it from usb ... if i do this the device will reset .. its strange ...
23:01:51amiconnIt's a nice player, but not the one with the really long runtime. The X5L has 2.5x the battery capacity of the plain X5, hence 2.5x the battery runtime, but only comes as single platter model (20GB and 30GB)
23:02:15TPM_Debianso i need a mount under linux ... to write the os ... but its my first time to mount a usb on linux .. and i dont know how i can made it
23:02:16 Join ATravelingGeek [0] (
23:03:35 Quit random_desu_is_s ("ChatZilla [Firefox]")
23:03:53 Quit iamben_ (Read error: 110 (Connection timed out))
23:04:05krazykitTPM_Debian, why exactly are you writing i2c.bin?
23:04:35Lloreanamiconn: I'm having slight lag issues, so I'm not sure when my message showed up. Why is AA going to be done without scaling?
23:04:46TPM_Debianwell ... my sansa will show black display on normal / recovery and manufactury mode ...
23:04:53amiconnLlorean: I don't know...
23:05:09linuxstb_Llorean: That's just the initial implementation Nico_P wants to commit. He's said he'll work on scaling afterwards.
23:05:10LloreanI can understand an argument for no jpeg, though I'm not sure where I personally stand on it.
23:05:27LloreanJust a starting point.
23:05:28TPM_Debiani can only access the device on manu-mode @ krazykit
23:05:31***Saving seen data "./dancer.seen"
23:05:34CtcpIgnored 2 channel CTCP requests in 27 seconds at the last flood
23:05:34*Bagder wants jpeg support and scaling for AA
23:05:47LloreanI'm indifferentish on jpeg, in favour of scaling
23:06:15amiconnHmm. I think a better solution would be using jpeg and scaling, but not putting the jpeg decoder in the core, but reusing the jpeg viewer plugin
23:06:29LloreanPlugins that return data to the core, then?
23:06:44amiconnNo, plugins that draw into a viewport...
23:07:08amiconnI *know* there's an ingredient missing to do that...
23:07:10DerPapstviewports that aren't implemented yet?
23:07:21LloreanWould that then conflict with TSR plugins?
23:07:26*DerPapst shuts up
23:07:36LloreanFor example, batterybench + AA/Jpeg WPS?
23:07:38krazykitTPM_Debian, so you can write the bootloader? after that, you can't mount the recovery mode, or do you not know how?
23:07:40amiconnLlorean: It would.
23:07:58linuxstb_Bagder: Could you update the gcc multilibs patch (in with
23:08:01amiconnBut it wouldn't conflict with normal plugin usage
23:08:17Lloreanamiconn: I didn't think it would, which is why I didn't mention them. :)
23:08:57TPM_Debiankrazykit, the only why, how i can "relive" the sansa is in the exacly steps ... 1. init manu-mode, 2. write bootloader, 3. hold rec befor device rebot abter write success, 4. display will show logo and report status ... 5. write os image ... but whisout mounting i cant write ... and when i disconnect usb after write bootload, all will erase ... (i think) ... no effect will be when i restart the device after disconnect (no display)
23:09:36Bagderlinuxstb: done!
23:10:02linuxstb_Bagder: Thanks. That _should_ now be fine for all current/in-progress ARM targets...
23:10:26krazykitTPM_Debian, so you're having problems mounting the device to write the OF image?
23:10:43krazykitdo you not know the command, or is the problem something else?
23:10:52BagderTPM_Debian: no, in manufacture mode you "reover" by loading the booloader so that you enter recovery mode, and then you should be able to mount the unit
23:10:55rasherLlorean: considering the kind of TSR plugins we have, I don't see that as a huge problem
23:11:08TPM_Debianits my first time to use a usb device under debina .. btw linux
23:11:23 Join n17ikh|Lappy [0] (
23:11:24TPM_DebianBagder, this will not work
23:11:24***Alert Mode OFF
23:11:40Bagderthen something is very wrong in your end
23:11:47Bagderhundreds of people have done that already
23:12:06TPM_Debianperhaps ... but my device will this not make
23:12:17 Join iamben [0] (n=ben@
23:12:31Bagderwhy not?
23:12:37Bagderyour explanations so far have been flawed
23:12:41Lloreanrasher: We only have one, right? But that one is exactly the sort you want to use with playback...
23:12:48TPM_Debianit will (i think so) destroy the bootloader after burn and power-off the device
23:12:52LloreanOr rather, with a view of the WPS
23:13:10TPM_Debianits very very strange
23:13:14rasherLlorean: But I'd say it's reasonable to say "don't use AA if you're running battery bench". We'll lose only the ability to specifically benchmark AA
23:13:20BagderTPM_Debian: can you explain what you want to do, and what exact device you have?
23:13:23rasherBut other than that, what difference does it make?
23:13:38Lloreanrasher: I would like to assume there might be future TSR plugins.
23:13:56LloreanI'm not saying we shouldn't do it this way, I'd just like to strongly encourage looking for another solution first.
23:14:59LloreanFor example, could we use a codec for the image, store the post-decoded/scale image in the metadata, since that also doesn't require viewports?
23:15:30linuxstb_Wow, the Creative Zen Vision:M New Ports forum thread has had 68880 views...
23:16:05LloreanWell isn't/wasn't that one of the top competitors to the iPod 5G?
23:17:35TPM_DebianBagder, "sansa e250" ... what the device not do: boot into normal-mode, boot into recovery mode ... what i do to "relive" ... 1. init manu-mode .. 2. flash bootloader out of debian .. 3. press REC befor write success and device will automatic reboot ... 4. device will wait for write os .... ... when i disconnect and reconnect usb on the device, display will be blank ... when the device reboot, and i do not push rec the device print "hold-key
23:17:35TPM_Debian is locked ... system shutdown" ... after manual power on the display will be blank .. in all modes ... so .. the only way i can flash the os is folowing the 4 steps, mount device and write os befor i disconnect the device from usb
23:18:03Dark_ApostropheIs Rockbox POSIX-compliant?
23:18:04amiconnLlorean: We have 2 tsr plugins...
23:18:09*Soap has spent a week trying to think of a cool TSR plugin - and has come up blank.
23:18:14Lloreanamiconn: What's the other one?
23:18:16amiconn(one is archos only though)
23:18:25LloreanAaaah, that's why I don't know of it.
23:18:30Soapperhaps his thoughts are too narrow.
23:18:55TPM_DebianBagder, or power will off
23:18:57BagderTPM_Debian: really, you should pay attention to details and report those.
23:19:06Bagderyour explanation is full of logical errors and flaws
23:19:15linuxstb_Dark_Apostrophe: Where it makes sense, functions in Rockboxare implemented to be POSIX-compliant. But you couldn't call Rockbox as a whole POSIX compliant.
23:19:22TPM_Debianmaybe ... my english is not perfect ;)
23:19:22*amiconn is going to try an adapted h1x0 radio driver on h10
23:19:57BagderTPM_Debian: your english is fine enough. Like in step two, there's absolutely nothing that flashes anything when your're in manufactrure mode
23:20:01LloreanThen again, TSR plugins that can draw to viewports are also a kinda nifty idea... so yeah.
23:20:10BagderTPM_Debian: and what happens exactly when it starts recovery mode?
23:20:15 Join kkurbju1 [0] (
23:20:16TPM_DebianBagder, nothing
23:20:24BagderI don't believe you
23:20:27 Nick kkurbju1 is now known as kkurbjun (
23:20:37Bagderso how do you know it started?
23:20:56amiconnLlorean: Mmmmyes, it would need to be tsr in that case. But I think that would be doable with the current infrastructure, just viewports are missing :/
23:21:17amiconnIf only more devs would care about the fundamentals - the low level code :\
23:21:34Lloreanamiconn: TSR plugins with viewports would then let people create "visualizations" in their WPSes, which they've only been asking for forever. ;)
23:21:35BagderI would rather say that AA through a plugin sounds like a hack
23:21:54rasheramiconn: I doubt it's a matter of care.
23:21:55Nico_Pamiconn: if only more devs knew about low-level stuff
23:21:58BagderI would rather then go for the full "plugin as WPS" concept
23:22:09amiconnIt would be rather flexible
23:22:20LloreanBagder: How about "WPS in plugin"?
23:22:42amiconnNico_P: That's something that can be learned
23:22:43Nico_Pamiconn: I wonder why I'd need to handle AA on lcd remotes
23:22:51 Nick parafin is now known as parafin|away (
23:23:03BagderNico_P: because people will want their WPSes on remotes to show them?
23:23:06LloreanBagder: Leave the current WPS as it is, but give plugins the draw the WPS (with their own additions) so that the more out-there WPS features could be in specific custom wps plugins.
23:23:11rasherNico_P: the m:robe has a rather large lcd remote iirc
23:23:25amiconnI knew next to nothing about low-level programming when first getting into touch with rockbox
23:23:26BagderLlorean: yes, that's pretty much what I think as well
23:23:34kkurbjunrasher: it's only 79x16
23:23:42Nico_Pthen the AA bitmap will be cropped for now. I don't want to drop a big bomb this time
23:23:42rasherAh, I was wrong.
23:24:01Nico_Poh well then who wants AA on the tiny screens the remotes have?
23:24:21LloreanNico_P: We've long since learned people aren't sane. ;)
23:24:24 Quit advcomp2019 (Read error: 113 (No route to host))
23:24:27linuxstb_Once the greyscale lib runs on the remotes...
23:24:29BagderNico_P: I'll bet they'll show their faces once you bring on the AA ;-)
23:24:39amiconnNico_P: It's not so much the different resolutions (btw. on M5 the main & remote lcd don't have very different resolution), but the different pixel formats
23:24:39 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
23:24:58Nico_Pamiconn: yeah that's not an easy one
23:25:25linuxstb_Doesn't it simply mean loading two bitmaps (or loading/scaling the same bitmap twice)?
23:26:04bertrikI thought the graphics API had the notion of a 'canonical' format? 24-bit for color, 8-bit for grayscale
23:26:48amiconnNot the core api, check the footnotes for those functions
23:26:52linuxstb_All bitmaps are stored internally in the native format for the LCD hardware.
23:27:00Lloreanlinuxstb_: I'd prefer scaling the original to two sizes, personally. I like the scale solution, especially if people have collections for use on multiple targets, or just multiple WPSes even.
23:27:03Nico_Plinuxstb_: yeah, but I think there are a few subtelties I'll need to work out
23:27:20amiconnHmm, looks like I get some communication with the radio chip :)
23:27:28Dark_ApostropheCan rockbox run any SDL stuff?
23:27:31przemhbHi all
23:27:36BagderDark_Apostrophe: nope
23:27:42Nico_PLlorean: for now I've implementing the search for different bitmaps, eg cover.100x100.bmp
23:28:34LloreanJust stating my personal preference. I'm not likely to ever use album art, anyway. :)
23:28:55*amiconn neither, other than for testing
23:29:10przemhbamiconn: have you tried someting new or tested the last patch?
23:29:17Nico_Pheh, you might actually come to like it ;)
23:29:35amiconnprzemhb: The patch I posted has some bugs of its own, and doesn't implement my latest idea
23:29:53amiconnBut it looks like I now have communication with the fm chip going
23:29:59LloreanNico_P: I do admit I have a little extra space in my WPS, but I'd like to fill it with enhanced playlist features eventually. :-P
23:30:24amiconn(and the patch won't apply to latest svn anyway because barrywardell did the target tree split)
23:30:42 Quit markun (Read error: 113 (No route to host))
23:31:06 Join markun [0] (n=markun@rockbox/developer/markun)
23:32:29Bagderkkurbjun: you did spot the red, right?
23:32:44przemhbamiconn: I've tried the latest Barry Wardell's patch and there is communication with radio chip
23:33:17kkurbjunbagder, thanks I'll fix that
23:33:31*amiconn never remembers those fs# numbers, and the search function in flyspray isn't that helpful either
23:34:02linuxstb_kkurbjun: I've now fixed the patch used by to correctly build libgcc.a for the arm1136jf-s if you want to test...
23:35:11amiconnprzemhb: Unfortunately barry isn't here now. Just omitting the clock stretch isn't going to work, the TEA needs it
23:35:26amiconnAnd I already suggested the solution - doing the same thing as on iriver
23:35:33amiconnThat's what I'm trying now
23:35:59kkurbjunlinuxstb, great, is it in svn?
23:36:01amiconnBut I need a full rebuild because the strings are all mixed up when only replacing the core
23:36:39linuxstb_kkurbjun: No, the patch itself is at - Bagder just updated it.
23:36:48TPM_DebianBagder, look this ... Http://
23:36:50amiconnAnd btw, for me booting didn't fail with your patch - that's because I've set a different start screen than main menu :)
23:36:55kkurbjun:), oh yeah
23:37:07kkurbjunI'll give it a try
23:37:16*amiconn is a long term rockbox user, hence *need* to have rockbox start up in the browser
23:37:23przemhbamiconn: lucky you
23:37:54amiconnIt's set this way on all my targets
23:38:21przemhbamiconn: my latest patch used H1x0 i2c code but it still wasn't enough
23:38:25BagderTPM_Debian: after step 5 when you've run recover, what did you see on the screen? I take it you then mounted the device? How did you "write new bootloader to device" ? what does step 7 "burn finish" mean?
23:38:32 Quit Arathis ("Bye, bye")
23:39:15BagderTPM_Debian: you could try to do the .fmt file creation to have the file system recreated
23:39:18amiconnprzemhb: Yeah, it's based on my patch, which contained 2 goofs...
23:39:35TPM_Debianstep 6 mean that the e200tool now writing the data to the device ... and step 7 says the Write cycle is successfully and e200tool exited
23:39:41amiconnReading SCL and SDA needs to read GPIOD_INPUT_VAL of course...
23:39:42przemhbamiconn: you're right
23:39:54LloreanBagder: Doesn't he need to hold record when it automatically reboots between 7 and 8?
23:40:03Bagderyes he does
23:40:14LloreanI thought it was running from RAM at that point, so if you let it shut down from the hold being on, as he says it does, you lose all progress.
23:40:28BagderTPM_Debian: see Llorean's comment, "recover" doesn't copy anything to any flash
23:40:53Bagderlike variant-2 does I see
23:41:04Bagderalthough that one seems to stop at step 10 for some reason
23:41:37TPM_Debiani hold rec only in variant 2 .. and the device will show the text in the recovery mode .. when i not do this all modes will show black screen after poweron ...
23:41:39BagderTPM_Debian: and when doing "recover" and holding REC, the system does not restart, it starts recover mode
23:41:55TPM_Debianvariant 2 is realy the only why that will print the recovery mode text
23:42:04Bagderyes, and you WANT the recovery mode
23:42:13LloreanTPM_Debian: In Variant 1 you're not IN recovery mode, you go back to manufacturing mode because you didn't hold Record.
23:42:14Bagderthat's the mode that you'll use to get things fixed
23:42:20LloreanYou never finished the recovery process.
23:42:38amiconnprzemhb: Seems like the tuner is receiving, just I don't hear anything yet
23:42:42 Quit jhulst (Connection timed out)
23:42:42przemhbamiconn: I've took a quick look at the radio chip readouts given by Barry and the readout I have on my own H10 - they look sensible to me
23:42:49amiconnAnd the radio screen buttons are messed up
23:43:00TPM_Debianlodesi, i think not so ... when the device will go back into mamu mode so i can access the device whis e200 .. but this will not works
23:43:18przemhbamiconn: turn the mute off and try to record sth from it
23:43:32przemhbamiconn: it may work
23:43:36amiconnBut from the stereo indicator it looks like I'm receiving the known stations in Berlin
23:43:38BagderTPM_Debian: really, if you don't follow the proper procedures you won't manage what you attempt to do
23:43:59LloreanTPM_Debian: Do variant 2. When you get text on the screen, attach it to a computer. Then copy an official MI4 file to it.
23:44:04Bagderbut feel free to ignore us
23:44:13amiconnNo buttons apart from up/down are working in the radio screen. Those 2 tune down/up
23:44:49przemhbamiconn: does statsion searching stops in proper positions (at your local radio stations frequencies)?
23:44:51amiconnI can't even leave the radio screen without shutting down or pulling the battery
23:45:29TPM_DebianBagder, sry .. i cant understand your text ... what u mean thiw "proper procedures" ?
23:45:47BagderTPM_Debian: go into recovery mode, mount the unit, copy the files
23:45:59LloreanTPM_Debian: He means you are not following the instructions.
23:46:08TPM_Debiana .. hmmmm
23:46:18amiconnprzemhb: I'll make a new patch, then you can experiment as well
23:46:29TPM_Debianwell .. how log i need to wait till the devie will go into manu-mode back?
23:46:30przemhbamiconn: great!
23:46:32amiconnJust need to adjust the config for big H10
23:46:43amiconn(did small H10 only as that's what I have)
23:46:43BagderTPM_Debian: ?
23:47:31TPM_Debianwhen the device will reboot after writing bootloader ... how long i need to wait till i can access the device whis e200tool ? ...
23:47:35BagderTPM_Debian: I get the feeling you're just playing with me, I'll stop this soon
23:47:46TPM_Debianno!!! ... realy no!!!!
23:48:11TPM_Debianthis is a real problem .. but my english is realy not the best!
23:48:27LloreanTPM_Debian: Then why aren't you doing what we tell you?
23:48:29 Join midkay [0] (n=midkay@rockbox/developer/midkay)
23:48:50LloreanDo variant #2, when you get text on the screen attach it to a computer, and copy the MI4 file to it.
23:48:51BagderTPM_Debian: then start it in manufacture mode, use e200too recover to go to recovery mode, then mount the system, copy the files, unmount
23:49:02TPM_Debianso i cant understand what all u type ... and googel translate can not all translate right .. its hard for me to understand you
23:49:23TPM_DebianBagder, how can i mount the device in debian?
23:49:23markunTPM_Debian: which languages do you speak?
23:49:35LloreanTPM_Debian: When it says "Welcome to recovery mode", attach it to your computer.
23:51:13TPM_DebianLlorean, after show "welcome" i cant disconnect usb .. if i do this the device will go off and i cant go into recovery .... so .. i need the available usb connection ...
23:51:31TPM_Debiani think, i need a manuel mounting of the device .. but i dont know, how i can made this
23:51:50LloreanTPM_Debian: When it shows "welcome" that IS Recovery
23:52:11TPM_Debianyes .. i know
23:52:22LloreanAll you need to do then is copy PP5022.MI4 to the player
23:52:31LloreanYou do NOT use e200tool with Recovery Mode.
23:52:32TPM_Debianyes .. but how?
23:52:32markunTPM_Debian: an you copied the MI4 file but you don't know how to unmount?
23:52:42Bagderhe doesn't know how to mount
23:52:53TPM_Debiani dont know how to mount
23:53:12DerPapstmount /dev/sdXN /mnt/bla?
23:53:17*DerPapst hides
23:53:24markunI guess you should learn that first or use a diffrent operating system
23:53:41TPM_Debianwhat is bla?
23:54:02Bagderand in recovery mode, it did come up as /dev/sdX to me, not sdXN
23:54:03DerPapsta foldername in /mnt :)
23:54:16*linuxstb_ is surprised Sandisk haven't released a recovery tool that works with manufacturing mode...
23:54:19LloreanTPM_Debian: You'll need to read the instructions for your operating system
23:54:21markunTPM_Debian: "bla" was just an example
23:54:21TPM_Debiana .. k .. i take a look into it
23:54:29DerPapstwell ipods use partitions so i was not sure ;)
23:54:32Lloreanlinuxstb_: Do we know the updater tool doesn't?
23:54:51BagderDerPapst: so does the sansa, although not in their recovery
23:54:54linuxstb_Llorean: No... But I assumed that if it did, people wouldn't be using e200tool
23:55:17DerPapstBagder: ah ok. didn't know that. but i'll learn once my sansa arrived
23:55:35TPM_Debianhmm ... i have no files in mnt ... ls -l will nothign print
23:55:44TPM_Debianthis cant be ?!?!?!
23:55:57Lloreanlinuxstb_: Well honestly, a "normal" user should never need a manufacturer mode tool.
23:56:22TPM_Debianwrong window ;)
23:56:39markunTPM_Debian: maybe ubuntu would be easier for you to use
23:56:44DerPapstTPM_Debian: where is your sansa? /dev/sd... ?
23:57:49amiconnprzemhb: Patch uploaded.

Previous day | Next day