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

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

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

#rockbox log for 2009-09-01

00:00:54 Join mgpaulus [0] (i=a62294f2@gateway/web/freenode/x-ejopoastaxofcsgb)
00:02:29mgpaulusHi all. Just installed Rockbox on my sansa e260R. Have it working fine with about 6GB of mp3. Just have one quick question that I didn't find an answer scanning for: What's the best way to "wake up" my player when the screen blanks, without messing with any settings, or going to some menu option that will mess with me?
00:02:54 Quit GeekShadow ("The cake is a lie !")
00:03:15JdGordon|enable the "first button press turns on backlight" setting
00:03:57mgpaulus<JdGordon> Thanks.
00:04:58*bertrik will write his dbop-button dev-mail tomorrow
00:05:44 Quit bmbl ("Bye!")
00:05:58 Quit mgpaulus ()
00:07:26 Quit stripwax_ ("")
00:09:05 Join Thundercloud [0] (
00:09:34kugelbertrik: I think you should wait until you have some hardware to test your proposals on
00:10:10TechGuyi thought about rockboxing my friends Sansa Fuze what are the risks of doing so?
00:10:17TechGuythe same as any mp3 player?
00:10:20bertrikI have some concerns that I can also voice without having the hardware I think
00:10:24 Quit webmind (Read error: 60 (Operation timed out))
00:11:02kugelI won't stop you :)
00:12:04 Join webmind [0] (
00:16:25kugelamiconn: can you tell me if there's a register where I could store 16bit in in lcd-as-e200v2-fuze.S?
00:17:03kugelmaybe some is known to be always only half full
00:17:22***Saving seen data "./dancer.seen"
00:21:38 Quit Zagor ("Clint excited")
00:24:15bertrikif I do 'make reconf' in a zenvision-m30 build dir, it errors out
00:24:37kugelarg, I can't get it to work with yuh+dithering
00:25:48 Quit jgarvey ("Leaving")
00:28:50 Quit bertrik ("De groeten")
00:35:57 Quit Newa (" boot")
00:39:57 Quit moos ("Rockbox rules the DAP world")
00:40:33 Join moos [0] (i=mostafa@rockbox/staff/moos)
00:44:30 Quit captainkwel ("Page closed")
00:45:25*kugel doesn't understand this
00:50:30 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
00:52:53kugelif anyone is able to see why lcd_write_yuv420_lines() works fine, but lcd_write_yuv420_lines_odither() not, please tell me :(
00:54:04 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-gbqljmrjpdihioho)
00:54:22saratogakugel: if hardware registers were read cached, it would be very hard to read from them
00:54:26saratogayou'd have to flush the cache on read
00:55:04kugelyea, that makes sense
00:56:48 Quit domonoky (Read error: 104 (Connection reset by peer))
00:57:00kugelsaratoga: you have e200v2, haven't you?
00:57:11saratogakugel: i do but am really busy today
00:57:20saratogawhat do you need
00:57:44kugeltesting of a patch here:
00:57:53kugelthe latest one is committed, so the one before
00:59:25 Quit PaulJam (Read error: 104 (Connection reset by peer))
01:02:15saratogakugel: hunk failed?
01:02:30saratogatheres only 1 hunk
01:02:52saratogahmm no there isn't
01:03:24 Quit Thundercloud (Remote closed the connection)
01:03:31saratogai have unpatched numbers, which are useful?
01:04:31kugelyou don't need the lcd-fuze.c hunk
01:04:35kugelthe others are needed
01:06:16kugelsaratoga: take this one
01:07:17saratogathat doesn't apply either
01:07:32kugelsvn up?
01:07:37saratogaalready did
01:07:39kugelI just did that and made the diff of it
01:08:27kugelthat gotta apply, I made it from a clean source
01:10:47saratogakugel sorry i have to go
01:10:47kugelsaratoga: this one maybe?
01:10:50 Quit saratoga ("Page closed")
01:13:04 Join polobricolo [0] (
01:24:09 Quit mcuelenaere ()
01:35:04 Join Strife89 [0] (
01:37:56 Join mc2739 [0] (
01:38:58 Part Necos
01:41:45 Join sean [0] (n=4083f4d3@gateway/web/cgi-irc/
01:41:45seanis this the Rockbox IRC?
01:41:52 Quit sean (Client Quit)
01:41:53tmztlooks like desktop linux stuff might be getting easier,
01:42:23tmztassuming rockbox exports something that could be used to identify that it's running
01:42:59mc2739kugel: tested next to last patch from FS #10272 against r22574 (would not apply to r22577) - test_fps unboosted: was 1/1: 44.7 fps now 1/1: 64.5 fps - test_fps boosted: was 1/1: 73.6 fps now 1/1: 100.0 fps
01:44:08JdGordon|so i dont need to test the e200v2?
01:44:18mc2739kugel: also fails to apply to r22577
01:45:34mc2739JdGordon|: not unless you really want to
01:46:49kugelthe very last patch?
01:47:03kugelHow does it not apply for you?! I create against clean svn :S
01:47:48kugelmc2739: alright, so please for boosting in the debug menu and look out if you see blue bars (or maybe small blue pixel)
01:47:56JdGordon|kugel: btw... I have no idea what your takling about with the fuze supposedly slow disk....
01:47:57mc2739lcd-e200v2.c - Hunk #1 FAILED at 173.
01:47:59JdGordon|its fine for me
01:48:06JdGordon|no slower than any other flash target
01:48:16kugelI never said it's slow
01:48:24JdGordon|yes you did
01:48:33kugelno I didn't
01:48:34JdGordon|you said it was unusably slow so you had to add dircache
01:48:54kugelI said it's not slow, but dircache still improves it noticeably
01:49:09JdGordon|same diff
01:49:19kugelnot the same
01:50:01 Join Tech-Guy [0] (
01:50:03 Quit Tech-Guy (Remote closed the connection)
01:50:05kugelbut you also apparently fail to see that dircache also improves battery life
01:50:11JdGordon|is the lcd bezel cutting off the top few pixels? or am i damn crazy?
01:51:05kugelthe case doesn't fit correctly on some/many fuze's. some of the top lines are hidden by the case
01:51:16JdGordon|thats what i mean
01:51:22*JdGordon| wonders if he is making up words again
01:51:46*JdGordon| isnt :)
01:52:50kugelthat was my whole point to finish costum ui vp finally :D
01:53:52*kugel has a new tiny, but probably still debatable, feature ready
01:54:10kugeldebatable due to the name of the setting of course
01:55:15*kugel waits for someone pretending to be interested
01:55:32kugelmc2739: do you notice anything weird?
01:56:03mc2739kugel: nothing weird boosted or unboosted
01:56:10kugelyou should be able to svn up after applying
01:56:26kugelapparently git is kidding me, I can't figure out why it doesn't apply for you
01:56:37JdGordon|kugel: uh oh! what?
01:57:46kugelJdGordon|: the new feature?
01:58:18kkurbjunamiconn: I'm running into another problem with the vertical packing.. the m:robe 500 lcd is effectively inverted from left to right (with vertical packing). I can store the images inverted and I could load them inverted, but if I were to use them inverted it would mean that I need to know that height stride and the actual width of the image if it is a _part call.
01:58:20kkurbjunIf I store the images non-inverted then the backdrop is inverted. Since just the left to right is inverted, but not the top to bottom I think I need to do divide/multiply's to figure out the backdrop location to make it appear correct for some of the pixel functions. I could also invert the backdrop only after it is loded, but that seems sloppy, I was wondering what you think the best approach would be for handling the inverted display.
01:59:20JdGordon|kugel: yeah
01:59:24kkurbjunI looked for the signals that are supposed to invert the display, but I don't think they are connected to the processor unfortunately
01:59:48kugelI love listening to gapless albums, so I always hate destroying the gap when skipping a track
01:59:53mc2739kugel: sorry, my mistake, I was still at r22574 - it applies fine to r22577 - building now
02:00:14kugelso just added a "jump to 5s before the track's end" option to skip length
02:00:41JdGordon|thats rather,... umm.... WTF?
02:00:53kugelwhy that?
02:01:10kugelWTF as in awesome? yea!
02:01:13JdGordon|thats just wierd :p
02:01:32kugelI can listen to the gaps that way
02:01:43kugelthe gapless gaps, that is
02:01:49JdGordon|of course!
02:01:55 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
02:03:28kugelkkurbjun: isn't inverting something you can get around with setting up the loops in lcd_update_rect with some magic?
02:03:47kugeland setting the window position
02:04:35kkurbjunkugel, not if I want to take advantage of the dma for lcd updates
02:04:38 Quit FlynDice (Read error: 131 (Connection reset by peer))
02:04:58kugelyea, understandable
02:06:03kkurbjunright now what I'm doing is writing a new lcd driver that has everything oriented as vertical stripes rather than horizontal stripes so that the rotation is taken care of at the lcd_ functions
02:06:27kkurbjunit really speeds things up, but the backdro pis a bit of a problem unfortunately
02:06:34kkurbjun200 fps versus 20 :-D
02:06:52kugelworth it!
02:07:20 Quit TechGuy (Read error: 110 (Connection timed out))
02:07:25kugeli've been speeding up some lcds today too, but 200fps is far away
02:07:27mc2739kugel: test_fps unboosted: 1/1: 100.0 fps - test_fps boosted: 1/1: 100.0 fps
02:07:44kugel200fps is an awful lot of data on the mrobe, isn't it?
02:07:56kugelmr500 that is
02:08:04kkurbjunyeah, 640x480x2 bytes
02:08:11kkurbjunper frame
02:08:57kugel123MB/s :p
02:09:16kugelthe AMS sansas cannot do more than 8 :/
02:09:27kkurbjunthe hardware dma takes advantage of bursting to/from the sdram so that helps quite a bit
02:09:37kugelwe're so much faster than the hardware
02:10:10kugelmc2739: nice, that's what I also get on my fuze. still no blue bars/pixel?
02:11:25 Join FlynDice [0] (
02:12:17 Join JdGordroid [0] (
02:12:45kugelmc2739: btw, is the scrollwheel any nicer with that patch? speeding up the lcd so much should vastly decrease the chance of getting blocked by it in our scrollwheel reading
02:12:48mc2739kugel: no blue bars or pixels
02:16:17mc2739kugel: scrollwheel seems much better
02:17:01 Quit JdGordroid (Remote closed the connection)
02:17:08kugelcompare it to pre-r22575 if you like
02:17:23***Saving seen data "./dancer.seen"
02:18:16mc2739kugel: I compared to clean r22574 and it is much smoother
02:18:30kugelcool! :)
02:19:04kugelthe fuze's wheel isn't so improved, since it works while lcd updates
02:25:44 Part froggyman
02:25:47 Quit andrewrb ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
02:25:53 Join froggyman [0] (
02:28:26 Join GreaterBeaver [0] (
02:28:36GreaterBeaveris it possible to play hdcd files on rockbox?
02:29:00GreaterBeaverwait i can
02:29:04 Part GreaterBeaver
02:30:36Mode"#rockbox +o moos " by ChanServ (ChanServ@services.)
02:30:50Mode"#rockbox -o moos " by ChanServ (ChanServ@services.)
02:31:02 Quit Torne (Read error: 60 (Operation timed out))
02:31:45kugelmc2739: going to commit it, thanks for testing
02:32:39mc2739kugel: you are welcome! I'm glad to help
02:39:16 Join Torne [0] (
02:42:27CIA-43New commit by kugel (r22578): Fuze/e200v2: Speed up the LCD further by letting the FIFO fill until it's full instead of keeping it almost empty. ...
02:43:24mooswell done !
02:44:32kugelthanks :)
02:44:54kugelUnhelpful: how much fps does the beast get in PF?
02:45:23kugelfuze has 55, and PF can't keep up with loading the covers :/
02:46:35Unhelpfulkugel: i said what i think the solution is. :)
02:47:26Unhelpfulinstead of threading loads, and letting the OS take care of scheduling them, do them them in the main thread, so we can control load-to-scroll ratio
02:49:28*Strife89 tries to strike up interest in FS #8806.
02:50:47CIA-43New commit by kkurbjun (r22579): Initial vertical stride LCD driver. Backdrop is inverted with it currently.
02:54:33kugelkkurbjun: isn't stride always the height in the new driver?
02:54:54kkurbjunnot exactly
02:55:00kkurbjunnot for _part call
02:55:05kugeljust wondering since the STRIDE macro is still in
02:55:30kkurbjunsince there are alot of images that are vertical strips the height is not always the stride
02:55:48kugelseems we also need a lcd-16bit-common.c :)
02:55:50kkurbjunyeah, the stride macro is required
02:56:06*JdGordon suspects there is a issue with line handling and is amazed it never caused a problem untill now :p
02:56:19kkurbjunalot of those common functions may have to change depending on how the backdrop needs to be handled
02:57:50CIA-43New commit by kkurbjun (r22580): Add in BMP loader for vertical stride and include lcd-16bit-vert.c in SOURCES
02:58:30JdGordonkkurbjun: all this is for fast lcd updating without really putting it into horiz mode?
02:59:20Strife89What are vertical strides, exactly?
02:59:38 Join FOAD_ [0] (
03:00:21kkurbjunJdGordon: yes, it makes it so that the lcd routines are oriented in the natural format of the lcd even though it is effectively rotated.
03:00:40 Quit FOAD (Read error: 60 (Operation timed out))
03:00:40 Nick FOAD_ is now known as FOAD (
03:01:18kkurbjunStrife89: I don't know if it's the right term, but instead of the pixels being oriented so that a sequential address increment moves horizontally, it moves vertically instead
03:01:20*kugel wonders if kkurbjun is planning on the ability to change orientation at runtime
03:01:46kkurbjunJdGordon: yeah, it makes the gui and the games much move responsive
03:02:00kugelof course, games are his motivation :p
03:02:04Strife89kkurbjun: Ah. :)
03:03:51kkurbjunkugel: unfortunately no, the images are oriented at compile time so it would take quite a bit to re-arrange them during runtime. I think there would be problems making LCD_WIDTH and LCD_HEIGHT variables too.
03:04:00*JdGordon now has bloxes where \n should be
03:04:41 Quit z35 (Read error: 54 (Connection reset by peer))
03:05:10 Join z35 [0] (
03:05:14Unhelpfulkugel: did you get fonts going nicely?
03:06:06JdGordonanyone game to try out a skin patch?
03:06:32Strife89JdGordon: What's it do, exactly?
03:06:41JdGordonhopefuly.... nothing noticable :)
03:07:34Strife89JdGordon: Sure, what's the FS#?
03:07:43kugelkkurbjun: oh :( You seem like the right man for the job :)
03:08:58Strife89JdGordon: I've got an iPod Photo and a Sansa c250.
03:09:04JdGordongogogo :)
03:09:52 Join langzeitstudent_ [0] (
03:10:07JdGordondont bother...
03:10:10JdGordondata abort on my ipod
03:10:33Strife89JdGordon: Already? :/
03:10:58Strife89Sheesh, I haven't even applied the patch yet.
03:12:07JdGordoni cant find the location in the map
03:12:35*Dhraakellian sees changes for the Fuze
03:12:44*Dhraakellian updates and starts building
03:13:15*Dhraakellian does somewhat miss being able to use precompiled binaries
03:13:24JdGordonand bloody hell.. it wont crash gdb
03:13:26kugelJdGordon: map is for looser, look into the disassembly directly
03:13:38Dhraakellianbut, well, I really like having a usable scratchpad button in Sudoku
03:13:41*JdGordon isnt that much of a masochist
03:14:10kugelthe map isn't very detailed. the disassembly shows you every address
03:15:04kugeland function and variable names
03:15:28Unhelpfulindeed. if you build debug binaries, you can use map or the linked binary to find the address offset from the function, and then the .o to look at the fault address in src+asm view :D
03:16:09kugelthe .o happens to be the whole rockbox.elf though, which makes a look into the map beforehand quite useless :)
03:16:47kugelwell, there are many .o's, but looking at rockbox.elf is very convinient :)
03:17:26Unhelpfulkugel: debug data seems to be stripped during our link stage :/
03:17:45Strife89JdGordon: Tell me if you get a new patch up. :)
03:17:53kugelno, stripping happens at the very end
03:18:14kugel.elf −− stripping −−> .bin −− scramble −−> .<target>
03:18:36JdGordonok, so how do i disssemble it to get the actual funciton?
03:18:41kugelrockbox.elf contains as many debug stuff as the .o's here
03:19:55 Quit Stephen_ ("Leaving")
03:20:07kugelJdGordon: do you know the function already?
03:20:53kugelarm-elf-objdump -D rockbox.elf > somefile.txt, then search the file for the address the data abort gave you. scroll then up for the function name
03:22:39JdGordonseems to be in clear_image_pos :<
03:23:24 Join fdinel [0] (
03:23:47JdGordonupdate_curr_subline definetly makes more sense...
03:23:56JdGordonnow how do i get the c line its in?
03:24:06kugelthat's the tricky part :)
03:24:23kugelalthough Unhelpful mentioned some src+arm view which I never heard off
03:24:36JdGordonits in one of the case: 's though
03:24:50*JdGordon has enough understanding of asm to figure that out :p
03:25:01Unhelpfulkugel: objdump -Sd <file> will mix asm and src.
03:25:06JdGordonstreq ?
03:25:13JdGordonstring equalt?
03:26:00kugelstreq means store if the eq flag is set
03:26:04 Quit jboy_ (Read error: 110 (Connection timed out))
03:26:07JdGordonok good
03:26:14JdGordonthere is no string compares in that func :p
03:26:15kugelthe address you want to store at is probably invalid
03:26:48JdGordonUnhelpful: file there being the .c or .o?
03:28:10kugelUnhelpful: that doesn't seem to work with arm-elf-objdump, it's not build with -g, is it?
03:28:48kugelJdGordon: is your curr_line pointer's maybe gone wild?
03:28:50JdGordon someone magic me :) the abort is at 00030bb0
03:28:59Unhelpfulkugel: it needs to be built with -g, and it needs to be the .o file, because the .elf has been through our linker script and had everything stripped
03:29:04JdGordonits almost certainly a null ref
03:29:22JdGordonits in the displayer though... not the parser
03:29:24kugelthat doesn't crash on arm as I noticed some days ago
03:29:45JdGordonwell.. a bad pointer of some sort
03:30:17JdGordonthe problem is in the parser though... cabbie loaded at boot works fine... cabbie loaded after another theme is chaos
03:30:23kugelarg, you got me to apply your patch
03:30:49Unhelpfulhowever, symbols are not assigned addresses yet in the .o, so you need to get the offset of the crash address from the symbol via the map or objdump on the .elf
03:31:42JdGordonldr     r2, [r1, #12] needs translating....
03:31:50JdGordonload register.... which?
03:32:06*JdGordon really should learn arm one day
03:32:29 Quit robin0800 (Read error: 110 (Connection timed out))
03:33:14*JdGordon adds a panicf() and checks the obvious pointer
03:33:40kugelJdGordon: "line->curr_subline->next == NULL" is the line
03:34:30JdGordoncurr_sibline is someone being NULL? which should be caught eariler?
03:34:32UnhelpfulJdGordon: load address-in-r1, +12, into r2. i think.
03:34:39kugel(I think at least)
03:35:04JdGordoneither way.. something is fucked :p
03:35:11*JdGordon gets going on dinner
03:35:49kugelJdGordon: it's definitely that lone
03:36:38Unhelpfulline could be NULL too... if you compile the source file in question with -fverbose-asm -save-temps you can get a .s file with comments saying which variables are in the registers.
03:40:35JdGordonits not... I just added a panicf() if !line and that got through
03:40:40kugelJdGordon: ah, sorry, it's " if (line->curr_subline->next)"
03:41:23kugelwhy did gcc put a beq to skip that str, instead of just strneq??
03:43:00*Strife89 heads off to bed.
03:43:02 Quit Strife89 ("If you hold a Unix shell to your ear, you can hear the C.")
03:45:09JdGordonoh goody... got gdb to segfault :)
03:45:25JdGordonp line
03:45:25JdGordon$1 = (struct wps_line *) 0x1
03:47:21 Quit froggyman ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
03:47:51*kugel wonders if gcc is smart enough to pad the chars in struct wps_line prorply
03:49:42kugelJdGordon: I know you want to save spaces, but the WPS_REFRESH_MASK should be an int IMO (or at least a short....)
03:50:03kugelI mean "unsigned char line_type;"
03:50:05JdGordontrying to keep changes minimal...
03:50:11JdGordonna, thats fine
03:50:28*JdGordon doesnt know if he fixed it or moved it :(
03:50:49kugelother parts are treating it as an int
03:53:30JdGordonthere is apparently a global variable line somewhere...
03:54:04JdGordonnope still crashes :(
03:58:20kugeli told you the line
03:58:24JdGordoncurr_line is somehow getting set to 1
03:58:30JdGordonyes gdb agrees with you
03:58:51*JdGordon really doesnt want to start this from scratch
03:59:04JdGordontucker time
04:01:47 Join timc [0] (n=aoeu@
04:04:22 Join CaptainKwel [0] (
04:05:30 Quit TheSeven (Nick collision from services.)
04:05:45 Join The_Seven [0] (
04:05:51 Nick The_Seven is now known as TheSeven (
04:06:36kkurbjunhmm, I can't access
04:06:37 Join Xerion_ [0] (
04:06:42CIA-43New commit by kugel (r22581): Move the FIFO emptying up, as having it after lcd_write_single_data16() causes problems with uneven transfers (I'm not sure why).
04:06:46kugelkkurbjun: I can :p
04:07:12kkurbjunahh, there it goes
04:07:20 Join TorneWuff [0] (
04:07:37CIA-43New commit by kkurbjun (r22582): Brickmania: Make the game speed scale with the GAMESCREEN_HEIGHT.
04:08:24kugelso you're having fun with fixing shitloads plugins?
04:08:31kugelI might have a job for you then :)
04:09:10kkurbjunhaha, no, I would rather be done with these - the plugins are a bit of a mess
04:11:59kugelthis macro is weird! "*GAMESCREEN_HEIGHT/176"
04:13:18 Join nirom [0] (
04:13:25kkurbjunyeah, it is a bit strange, but I was not sure when the compiler did the division so I wanted to make sure that the order of operations was carried out: mult then divide
04:14:42kugel6 SPEED_SCALE or 6*SPEED_SCALE is totally the same, and doesn't have anything todo with the compiler
04:14:48 Quit Torne (Read error: 111 (Connection refused))
04:15:36kkurbjunif speed scale was just defined as GAMESCREENHEIGHT/176 is it possible for the compile to do that division before substituting it in for SPEED_SCALE?
04:16:02kugelthe compiler doesnt see that at all
04:16:08kugelit's all preprocessor
04:16:42kkurbjunahh, so the preprocessor does the substitutions and then the compiler calculates anything that is static?
04:17:07kkurbjungotcha, I'll change that define then
04:17:24***Saving seen data "./dancer.seen"
04:17:26kugelyou can make sure of the order by using parenthesis
04:18:38kkurbjunyeah, order of operations should be fine if it's substituted before evaluated
04:20:18kugelbubbles is plenty fast with 100fps
04:20:33kugellooks a bit turbo to me :p
04:24:05 Quit Xerion (Connection timed out)
04:24:05 Nick Xerion_ is now known as Xerion (
04:24:50 Nick nirom is now known as nirom_away (
04:25:02 Quit Lss (Read error: 104 (Connection reset by peer))
04:28:37Unhelpfulthe compiler should fold pretty much any subexpression that has only constant inputs into a single constant value
04:30:04 Quit CaptainKwel (
04:30:04 Quit J-23 (
04:30:04 Quit goffa__ (
04:30:04 Quit SUSaiyan (
04:30:04 Quit rwong (
04:30:04 Quit shaggy-h (
04:30:04 Quit cg_ (
04:36:02NJoinCaptainKwel [0] (
04:36:02NJoinJ-23 [0] (
04:36:02NJoingoffa__ [0] (n=goffa@
04:36:02NJoinSUSaiyan [0] (
04:36:02NJoinrwong [0] (
04:36:02NJoinshaggy-h [0] (
04:36:02NJoincg_ [0] (
04:38:42 Join dys` [0] (n=andreas@
04:40:20CIA-43New commit by kkurbjun (r22583): Brickmania: Fix define and add support for vertical strides.
04:41:00 Nick nirom_away is now known as nirom (
04:44:14CIA-43New commit by kkurbjun (r22584): Star: Add support for Vertical strides.
04:46:45 Quit rvvs89 (Read error: 60 (Operation timed out))
04:48:10 Quit Rondom (Nick collision from services.)
04:48:20 Join Rondom [0] (
04:49:52 Quit mc2739 ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
04:50:03 Quit timc (Read error: 104 (Connection reset by peer))
04:51:08 Quit dys (No route to host)
04:53:41 Quit nirom (Remote closed the connection)
05:07:29 Quit kugel (Read error: 110 (Connection timed out))
05:10:03 Quit JdGordon ("Leaving.")
05:10:27 Join rvvs89 [0] (
05:21:01 Join bzed_ [0] (
05:21:01 Quit bzed (Read error: 104 (Connection reset by peer))
05:21:18 Nick bzed_ is now known as bzed (
05:23:10CIA-43New commit by kkurbjun (r22585): Invadrox: Add support for vertical strides
05:40:26CIA-43New commit by kkurbjun (r22586): Matrix: Add support for vertical strides
05:42:24 Quit Horscht ("Verlassend")
05:44:07 Join n1s [0] (n=n1s@rockbox/developer/n1s)
05:49:42 Join tvelocity[a] [0] (
05:50:26 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
05:53:44 Quit tvelocity (Read error: 110 (Connection timed out))
05:55:48 Part jordan`
05:56:40 Join jordan` [0] (i=gromit@
05:58:41 Join timc [0] (n=aoeu@
06:02:24 Quit timc (Read error: 104 (Connection reset by peer))
06:03:52 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
06:07:52 Join timc [0] (n=aoeu@
06:14:09 Join toffe82 [0] (n=chatzill@
06:17:28***Saving seen data "./dancer.seen"
06:29:39 Join JdGordroid [0] (
06:29:50 Quit timc (Read error: 110 (Connection timed out))
06:31:40 Join timc [0] (n=aoeu@
06:32:08 Quit panni_ ("( :: NoNameScript 3.81 :: )")
06:39:52CIA-43New commit by kkurbjun (r22587): Add support for vertical stride to resizer.
06:40:57 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
06:42:38CIA-43New commit by lowlight (r22588): Enable sound on the Samsung YH-920
06:57:10 Join AndyIL [0] (n=pasha_in@
07:00:17 Quit AndyI (Read error: 110 (Connection timed out))
07:07:56 Quit Llorean ("Leaving.")
07:19:17 Join AndyI [0] (n=pasha_in@
07:20:03 Quit AndyIL (Read error: 60 (Operation timed out))
07:44:32 Quit CaptainKwel (Remote closed the connection)
07:48:22 Quit timc (Read error: 60 (Operation timed out))
07:52:28 Join l403 [0] (n=l@
07:55:48 Part toffe82
07:57:09l403I have looked at this list and checked the first task, to organize the settings list, the list seems to be organized already so the task can be removed from the list
08:02:16JdGordroidnot at all
08:02:28JdGordroidits still a huge mess
08:05:10l403ok, checking again
08:05:32JdGordroidits mostly in generation groups unit its far from what wed call organised
08:07:29 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
08:16:26l403I have found two closing parentheses that dont have their opening. they are highlighted in red in vim on line 600 and 1303 settings_list.c
08:17:29***Saving seen data "./dancer.seen"
08:19:00JdGordonyour line numbers are wrong...
08:19:06JdGordon600 is a #else
08:20:36 Join pamaury [0] (
08:21:07pamaurypcc1: hello, funman told me yesterday that you were interested in MTP for rockbox
08:22:07 Join intrados_ [0] (
08:22:15l403 /* display */ > CHOICE_SETTING > #else > ID2P(LANG_INVERT_CURSOR_BAR))
08:22:46 Quit intrados (Read error: 104 (Connection reset by peer))
08:23:37JdGordonthe next character after CHOICE_SETTING is the (
08:24:09 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
08:25:10l403but that is closed after ID2P(LANG_INVERT_CURSOR_GRADIENT))
08:25:46 Quit JdGordroid (Remote closed the connection)
08:27:21amiconnkkurbjun: Maybe I don't understand the backwards problem fully, but I think you have two options:
08:28:14amiconn(1) Make the bitmap format even weirder, so that the images go right-to-left like the framebuffer
08:29:01amiconn(2) Flip everything to left-to-right, and change the lcd update to go column by column. You need that for partial updates anyway
08:29:32l403I guess I just need to study more
08:29:57amiconnI also think that the lcd_framebuffer definition needs changing. The LCDADDR() macro looks horrible...
08:29:58 Join flydutch [0] (
08:30:15amiconnI mean it needs changing for this special pixel format
08:31:50 Quit pamaury ("Parti")
08:31:50amiconnAnother option would be to give up on this, go back to using plain lcd-16bit.c, and optimise the update function . It would mean to render quite some work useless though :\
08:33:47amiconnI also wonder whether the lcd controller allows to set the update direction (I mean horizontal <-> vertical). Many lcd controllers have that feature
08:35:46 Quit AndyI (Read error: 145 (Connection timed out))
08:35:50 Join AndyI [0] (n=pasha_in@
08:37:04 Join Rob2223 [0] (
08:37:26 Nick polobricolo is now known as polobricolo_ (
08:38:41 Quit safetydan ("Leaving.")
08:40:12 Join ender` [0] (
08:45:24 Quit Rob2222 (Read error: 145 (Connection timed out))
08:46:08 Join pamaury [0] (n=pamaury@
08:49:15 Join AndyIL [0] (n=pasha_in@
08:49:46 Quit Zagor ("Don't panic")
08:55:59 Join timc [0] (n=aoeu@
08:57:21 Quit webmind (Read error: 60 (Operation timed out))
08:57:47 Join webmind [0] (
08:58:03 Quit BHSPitMonkey (Remote closed the connection)
09:00:23 Quit AndyI (Read error: 110 (Connection timed out))
09:06:48 Join funman [0] (n=fun@rockbox/developer/funman)
09:11:56 Quit l403 (Read error: 113 (No route to host))
09:16:48 Join einhirn [0] (
09:18:51 Join l403 [0] (n=l@
09:18:57JdGordoncan someone please have a look at FS #10576 and see if you can spot bad code? somewhere curr_line (I tihnk) is getting corrupted, and I'm 99% sure its in skin_parser.c but for the life of me I cant see it...
09:19:18JdGordonit seems to only happen after loading some themes after other ones...
09:19:25 Quit l403 (Client Quit)
09:19:40JdGordone.g on the video cabbie works fine loaded from boot, but not after some other theme is loaded first
09:20:43funmanJdGordon: where is the crash?
09:21:24JdGordonthe actual crash happens in update_curr_subline() in skin_display.. but by the the line is completly corrupted
09:22:25funmandid you try running the sim in valgrind?
09:24:01JdGordonnope, ill give that a quick go
09:25:44JdGordonnup, nothing in valgrinds output out of the ordinary... and I did get to my panicf() so it did happen
09:26:46funmanno panicf in
09:27:06JdGordonno, i added it locally...
09:27:24JdGordonits at the start of of update_curr_subline() if line==1
09:28:03funmanbut line is a pointer?
09:28:15JdGordonline is somehow getting set to 1!
09:28:36funmanoh ok, can you see what did set it to 1 in the backtrace ?
09:29:26JdGordonhmm/... going by some debug output it looks like its happening pretty late in the redraw... maybe the last line of the skin
09:30:21funmanhow can i trigger the bug ?
09:30:27 Join Thundercloud [0] (
09:30:48JdGordonipod video sim, load the boxes theme, then load the cabbiev2 theme, then start music
09:30:57JdGordonor return to music
09:32:39JdGordon0x790001 doesnt look like a correct pointer... does it?
09:32:53JdGordonthey should all be 32bit aligned
09:33:22funmanif you do 32bits access yes
09:35:34funmancurr_vp->lines = 0x9bd in parse_progressbar
09:36:17 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
09:39:18funmanskin_data_load() doesn't initialize lines ?
09:40:01funmanah with wps_start_new_line() ..
09:40:59JdGordonno, I tihnk thats it
09:41:12JdGordoncurr_vp->lines needs to be set to NULL before going into that the first time
09:41:37JdGordondamn, nope
09:42:53funmanIn 2 crashs I see skin_viewport->lines (skin_display.c) and curr_vp->lines (skin_parser.c) are corrupted
09:43:39JdGordonalmost certainly caused by the parser though
09:43:49JdGordonbut yes, thats the same I'm seeing
09:45:37JdGordonSON OF A BITCH!~
09:45:44JdGordonfunman: you win the internet!
09:45:57JdGordonvp->lines isnt being initlaised in parse_viewport()
09:46:49funmanhm but what does wps_start_new_line do ?
09:47:11JdGordonit adds a new line to the current viewports list of lines
09:47:30funmanah so it should be set to null before calling this
09:48:11funmancool, 1 internet ^^
09:48:55JdGordonnow to test on target and update the patch.. and hope it gets some testing overnight :)
09:48:58JdGordonthanks alot
09:52:24JdGordongrr.. now a different theme doesnt work
09:52:35JdGordonbut now its bed time
09:55:20pamaurypcc1: tell me when you are there
09:57:51 Quit Thundercloud (Remote closed the connection)
09:58:02CIA-43New commit by mcuelenaere (r22589): Onda VX747/VX777 bootloader: power off after holding the poweroff button for 2 seconds (BUTTON_REPEAT doesn't seem to work in bootloader)
10:00:15pcc1pamaury: I'm here
10:00:56pamaurypcc1: funman told me yesterday that you were interested in MTP for rockbox
10:03:25pcc1pamaury: yes. I wanted to know if your work in progress was available anywhere so I could possibly help
10:04:48pamauryCurrently it's only on my computer but that can changed :) Have you read the MTP spec / have you got some usb knownledge in rockbox ?
10:06:59pcc1I scanned the spec but I have no other experience. I generally learn as I implement when doing this sort of development
10:08:55pamauryOk. Have you got a svn repo or something to share the code ?
10:11:44pcc1have you used git before?
10:12:22pcc1the code can be shared on github or something
10:12:23pamauryNo but I don't mind, it should be easy to learn to do simple things
10:14:11CIA-43New commit by rasher (r22590): Add the Onda 777 to the simulators list.
10:14:26pamauryOk, I'll have a look at github. Can we talk in private messages, that would avoid spamming the whole channel ?
10:17:15GodEaterpamaury: please don't
10:17:28GodEaterit's not spam when it's about Rockbox
10:17:34***Saving seen data "./dancer.seen"
10:17:52GodEaterother developers would definitely be interested in reading the conversation
10:18:01pamauryThat just for the git things
10:18:15pamauryNot sure developpers are interested in the setup of a git repo :)
10:18:23GodEateryou might get some more help :)
10:18:33GodEaterthere are a reasonable number of git fans here
10:19:24rasherI'd say git talk belongs in #rockbox-community
10:20:51pcc1in PM I asked pamaury to get the latest git repo following apply your patch, upload it and let me know the URL
10:21:41pamauryI'll setup something somewhere and come back when it's ready
10:22:07pcc1has anyone actually tried uploading the repo to github? I saw a 100MB limit, but my .git directory is about 110MB
10:23:10pamauryI see a 300MB limit on github
10:23:43pcc1hm, there were conflicts between the google search result and github's own page. I guess the latter overrules the former :)
10:24:40GodEaterI think I actually set the rockbox repo up on github way back when is another good choice, I don't think they have a size limit.
10:24:55GodEaterat least, I've never hit one.
10:26:02GodEaterhehe - indeed it is there. Needs an update though - I last pushed to it in 2008 ;)
10:26:23pamaurywhat is the url ?
10:28:55GodEaterI think he meant my github copy
10:29:34GodEaterI need to work out how to add it as a remote to my local repo though
10:29:38*GodEater reads git docs
10:30:07pamauryDoesn't matter. If you have a git repo that you don't use and that we can use for our MTP tests
10:32:28funmanGodEater: git-remote ?
10:34:36GodEaterfunman: done
10:34:40GodEaterI'm just doing the git push now
10:35:05pamaurySo this repo will be up-to-date ?
10:36:01pamauryAnd what can I do with it ? Should I fork it ?
10:36:29GodEaterI can add you as a collaborator on it
10:37:00pamaurySo I can modify it freely to work on MTP ?
10:37:05GodEatersure - just make a branch for MTP on it
10:37:23pcc1pamaury: or you can clone it to create your own repo. git being a DVCS, each developer generally has their own repo
10:37:26*GodEater 's connection is noticeably slower whilst the push happens
10:39:13pamauryThe best way for me would be to avoid to duplicate my modified svn local copy of RB on my computer but perhaps I'm asking too much
10:41:01pcc1hm, having two VCSs coexist in the same repo seems like trouble. you should be able to create a patch with "svn diff" and apply to the git repo
10:41:27GodEateryeah, you could always use git-svn with the main rockbox repo too
10:41:33GodEaterand just add the github one as a remote
10:41:49GodEaterlooks like the push is done
10:42:40*GodEater wonders how long it takes github's web interface to update
10:43:08pamauryMy knowledge of git is poor, what does it changes to have github being a "remote" of my local repo as well as the main rockbox repo ?
10:43:49GodEaterpamaury: it's not so much a git concept as a DVCS concept in general
10:44:04GodEateryou can pull(or push/commit) code from multiple places
10:45:04pamauryOk, so I will first use git-svn with the main repo like describe in the wiki and then add the gibhub
10:45:14GodEaterthat would be the ideal way ;)
10:45:21GodEaterjust make sure you take an svn-diff first
10:45:38GodEaterso you can apply your current changes to your git repo once it's set up
10:46:19pamauryWhat command should I run ? I can just convert my svn repo to a git-svn ?
10:47:11 Quit langzeitstudent_ (Remote closed the connection)
10:48:51pcc1pamaury: there's no automated process. once you have a git repo you can create a patch against svn trunk with "svn diff" and apply it to the git using "patch"
10:51:27GodEaterthe github repo should now be up to date with svn
10:51:51rasherIsn't there a proper git mirror these days?
10:52:02GodEateryes, but not one non-commiters can push to
10:52:44GodEaterhe can add that as a remote too :)
10:57:12rasherWhat's the point of all these plugin that splash their name when you start them?
10:58:49 Join jgarvey [0] (
10:59:09 Quit timc (Read error: 54 (Connection reset by peer))
11:00:02 Nick polobricolo_ is now known as polobricolo (
11:01:19pamauryThat's done. How do I add the gib-hub as a remote ? git remote add ???
11:04:35GodEaterI think you need a github id
11:05:04pamauryWhich means ?
11:05:11GodEaterhave you signed up there ?
11:05:24GodEaterright, so what's your id ? :)
11:05:38 Join timc [0] (n=aoeu@
11:05:55pamaurylogin ? pamaury
11:06:13GodEaterI've added you as a collaborator
11:06:18GodEateryou should see the Rockbox repo in your list now
11:06:27GodEaterand it should give you "Your clone url"
11:06:40GodEaterwhich you use as "git remote add github <your clone url>"
11:08:07pamauryOk that's done. So I should now branch to work on MTP, no ?
11:08:18 Join BdN3504 [0] (n=55b2236f@gateway/web/cgi-irc/
11:09:02pcc1pamaury: yes. man git-branch
11:09:08BdN3504which targets do NOT support the repeat a-b mode?
11:09:50BdN3504it's not listed in in the features.txt file, that's why i ask
11:13:23pamauryI'm not familiar with branching, once I have created a branch, should I specify which branch I'm working one ?
11:13:39 Join robin0800 [0] (
11:14:11GodEateryou do "git branch <my-new-branch-name>"
11:14:15pcc1once you have created the branch, use "git checkout <name-of-branch>" to switch to that branch
11:14:21GodEaterand then "git checkout <my-new-branch-name>"
11:14:31GodEaterand then all your commits will be on that branch
11:15:40GodEaterpcc1: do you have a github ID I can add to this repo too ?
11:16:39pamauryOk, I read that git was a decentralized repo, contrary to svn, but I guess there is a nothing of commit ? git push will commit changes on github ?
11:17:04GodEaterprovided you tell git which remote to push to, yes.
11:17:11 Quit BdN3504 ("CGI:IRC (EOF)")
11:17:35GodEaterso provided you called the github remote "github" as I indicated earlier in the command I gave you to add it, then "git push github" will commit to the repo there.
11:17:59pcc1GodEater: no. I was planning on hosting my changes outside of github
11:18:58GodEaterpcc1: cool
11:19:44 Join PaulJam [0] (
11:20:28pamaurypcc1: if you don't host on github, then where will you host them ?
11:22:03pcc1I will host on a VPS which I rent. you will be able to add it as a remote to get my changes
11:22:32 Quit xavieran (Read error: 104 (Connection reset by peer))
11:23:43pamauryAnd will I be able to do changes on that repo ? It can be useful if you want to work with me on MTP :)
11:24:24pcc1I can pull your changes from the github repo and add them to mine
11:25:14 Join decayedcell [0] (
11:25:54pamauryI have to leave now. I'll come back later to discuss the detail MTP.
11:26:03 Quit pamaury ("exit(*(int *)0 / 0);")
11:35:38 Nick funman is now known as away (n=fun@rockbox/developer/funman)
11:35:48 Nick away is now known as funman (n=fun@rockbox/developer/funman)
11:36:56 Quit Sajber^ (Read error: 104 (Connection reset by peer))
11:40:39CIA-43New commit by lenzone10 (r22591): Added a missing line for spoken QUICKSCREEN_TOP item in Quick Screen.
11:46:28 Join kugel [0] (n=kugel@rockbox/developer/kugel)
11:48:07 Quit funman ("free(random());")
11:51:26 Join funman [0] (n=fun@rockbox/developer/funman)
11:56:05 Quit moos (Read error: 131 (Connection reset by peer))
11:56:12 Join moos [0] (i=mostafa@rockbox/staff/moos)
11:59:00 Join petur [50] (n=petur@rockbox/developer/petur)
12:17:36***Saving seen data "./dancer.seen"
12:18:05 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
12:21:48 Join _zic [0] (
12:21:48 Quit kugel (Read error: 110 (Connection timed out))
12:29:32 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
12:30:28 Quit moos ("Rockbox rules the DAP world")
12:30:46 Join moos [0] (i=mostafa@rockbox/staff/moos)
12:34:02 Nick polobricolo is now known as polobricolo_ (
12:44:14 Quit funman ("leaving")
12:53:30 Join VoltageX [0] (
12:53:56VoltageXhi, can I use rockboy on the sandisk e260?
12:54:08 Quit polobricolo_ (Read error: 110 (Connection timed out))
12:55:49 Part webmind
12:57:35 Part decayedcell
12:57:58 Join decayedcell [0] (
12:59:02krazykitVoltageX, as detailed in the manual:
13:00:20VoltageXoh! it's not a distinct plugin any more
13:00:26VoltageXfile viewer now.
13:01:19gevaertsthat's still a plugin
13:06:40 Join AndyI [0] (n=pasha_in@
13:08:09 Join pamaury [0] (n=pamaury@
13:09:30VoltageXgevaerts: ok.
13:10:09 Join Creposucre [0] (n=86f69562@gateway/web/cgi-irc/
13:11:01CreposucreHi, is it possible to edit the title of a patch? How can I do that?
13:16:16 Join FOAD_ [0] (
13:17:41 Quit AndyIL (Read error: 110 (Connection timed out))
13:18:42PaulJamCreposucre: do you mean on Flyspray? i think only the devs can edit the titles of FS tasks.
13:20:54CreposucreYes, on flyspray. I would like to edit my task (10494) to make it more complete
13:33:08 Quit FOAD (Read error: 110 (Connection timed out))
13:33:08 Nick FOAD_ is now known as FOAD (
13:34:41 Join Llorean [0] (n=DarkkOne@
13:36:47AlexPCreposucre: A related question on your patch - how portable is the RDS part of it? Would it be fairly easy to use that for other targets that have RDS, e.g. the gigabeat S?
13:37:37AlexPCreposucre: Also, I can edit the task for you if you want - what would you like changed?
13:42:09Zagorwhere is __PCTOOL__ defined?
13:42:39Zagorah, in the Makefiles
13:55:35 Join panni_ [0] (
13:56:45 Join chandoo [0] (
14:02:12 Join mt [0] (n=MTee@rockbox/developer/mt)
14:04:26 Quit petur ("later")
14:04:26 Quit Llorean ("Leaving.")
14:11:24CreposucreAlexP: I removed the RDS part for the moment, but it was basically displaying two lines with RDS info; thoses infos were "computed" by the driver part
14:11:56CreposucreAlexP: so it's not really an "RDS architecture"
14:13:04CreposucreAlexP: about the FS, i'd like to add in the title "+ improved accessory detection" to attract people interested in thoses
14:17:37***Saving seen data "./dancer.seen"
14:18:06CreposucreAlexP: back to the RDS stuff, if the last patch works, I'll try to re-add thoses RDS info lines with some ifdef
14:39:30 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
14:40:34 Join chandoo_ [0] (
14:44:34 Join chandoo__ [0] (
14:46:10 Quit w1ll14m_ (Read error: 60 (Operation timed out))
14:47:45 Join funman [0] (n=fun@rockbox/developer/funman)
14:50:57dionoeaCreposucre: about the fm remote patch: I haven't had time to give it a try again. I'll do that tonight. (I'm Antoine on the tracker)
14:51:25dionoeaI'll use a clean rockbox tree to make sure that I don't have any potentially conflicting changes
14:52:09funman(I'm Antoine on the illegal bittorrent tracker)
14:52:22funmanhm sorry for that
14:56:40 Quit chandoo (Read error: 110 (Connection timed out))
14:58:14 Nick fxb__ is now known as fxb (
15:00:17 Quit chandoo_ (Read error: 110 (Connection timed out))
15:02:59 Join l403 [0] (n=l@
15:07:14Creposucredionoea: ok, no problem ;)
15:18:35 Quit mt (Read error: 104 (Connection reset by peer))
15:22:32 Quit _zic (Remote closed the connection)
15:22:40 Join chandoo_ [0] (
15:25:27 Quit PaulJam (Read error: 113 (No route to host))
15:27:58 Join chandoo [0] (
15:38:01 Quit bluebrother (Nick collision from services.)
15:38:06 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
15:38:20 Join helfinpzy [0] (n=51ecf22e@gateway/web/cgi-irc/
15:38:32 Quit chandoo__ (Read error: 110 (Connection timed out))
15:41:04 Quit VoltageX ()
15:44:01 Quit chandoo_ (Read error: 110 (Connection timed out))
15:45:27 Quit helfinpzy ("CGI:IRC (EOF)")
16:05:33 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-hepufmokpwyuwbxj)
16:05:41 Part decayedcell
16:05:47saratogakugel: are you going to post bootloaders for the YH players?
16:06:10funmanmy yh920 just went crazy
16:06:29funmancorrupted audio output, screen was locked, fat corruption when i restarted
16:06:36funmani believe because battery was empty
16:07:05saratogathe 920 is the only one with voltage monitoring right?
16:07:11 Join kugel [0] (n=kugel@rockbox/developer/kugel)
16:07:38funmanno it's the 925
16:08:16saratogahmm so no low voltage shutdown
16:08:28saratogaare there scans of the 920 somewhere?
16:08:32 Join Sajber^ [0] (
16:08:32saratogadoes it use the same chip as the others?
16:08:49funmanit's the same PP i think (is it what you mean?)
16:09:28funmanlow res scans from pyro_maniac
16:09:46funmani could scan them if needed, just have to unmount it
16:10:42 Join TopyMobile [0] (
16:12:54 Part LinusN
16:13:15saratogahuh the charger on the YH920 doesn't monitor the voltage aside from checking for low battery
16:13:33saratogaso i guess voltage monitoring must be done directly via the PP chip
16:13:52saratogai wonder if the code for one of the other PP chips could be reused
16:17:41***Saving seen data "./dancer.seen"
16:19:13saratogafunman: you could try calling adc_read as the H10 does
16:19:25saratogathey use nearly the same charge controller, and the same CPU
16:19:44saratogai think the same driver should work on both, though its possible its on a different ADC channel
16:21:37funmansaratoga: the h10 code is present in powermgmt-yh920.c but commented out
16:24:16funmani see a battery read out, let's see if it discharges
16:24:21 Quit l403 (No route to host)
16:25:41 Join kkurbjunW [0] (n=karlk@
16:27:13kkurbjunWamiconn: the problem with inverting the image is that if someone makes a _part call with say a source x offset of 4 and a width of 5, but the image is actually 10 pixels wide, I cannot properly calculate the source starting point without with full width and height of the image
16:28:21kkurbjunWI could just invert the backdrop, but I was a little concerned about that since it would mean there are multiple formats stored in memory. It shouldn't be a problem, but I don't know how or if the backdrop handling will change in the future.
16:29:36funmansaratoga: it works fine, now i'm not sure how to calibrate the battery levels
16:30:05amiconnkkurbjun: What about the other ideas?
16:31:16 Quit TopyMobile (Read error: 110 (Connection timed out))
16:32:05saratogafunman: enable voltage reading and then do a battery bench
16:32:21saratogathe lookup table values can be generated from the battery bench curve
16:33:50saratogaon their first day supported, the YH players are going to be the most stable port ever
16:33:59saratogathey'll have everything except maybe recording
16:34:35funmani'll let it charge for a bit
16:34:59funmanyh920 has some lcd glitches in splash screens
16:36:04 Join tvelocity[away] [0] (n=tony@
16:37:00CIA-43New commit by funman (r22592): YH920 : enable battery reading (uncalibrated yet)
16:38:48funmanhm the YH92x wiki mentions radio for yh920, i don't remember if mine has a FM tuner
16:39:34 Join tvelocity [0] (
16:39:35funmansaratoga: they'll only miss the manual
16:41:26 Join Lss [0] (
16:42:43 Join chandoo_ [0] (
16:47:49kkurbjunWamiconn: well, I wouldn't want to go back on the work.. changing the updates to go column by column might be possible, I need to think about that more. Right now it is doing that, but it's in blocks of 32x32 pixels. I think I could change that to strips, but it would likely have an effect on performance.
16:47:50kkurbjunW the LCD does have an option to reverse the updates, but it is supposed to just be controlled by a high/low signal similar to the VGA/QVGA mode it has. I have not found a pin that controls the update direction unfortunately, so my suspicion is that the designers grounded/drove the pin high direct with a power rail (or a pullup/pulldown)
16:49:12kkurbjunWI think updating the screen in strips could be the most promising. I need to see how many extra bursts that would take if I was just doing it by columns. If the width of the copy is not a multiple of the burst the copy still takes the same amount of time that an extra burst would
16:49:32kkurbjunWso if you are one over a burst multiple it will take a full extra burst cycle
16:49:57saratogadoes anyone have a YH820 that could test battery reading on it?
16:50:12 Quit saratoga ("Page closed")
16:51:01kkurbjunWhmm, I think that the width of the lcd will work out so that it is exactly on the burst size - the processor will be interrupted to start the copies a bit more often, but I bet the performance impact wouldn't be too bad..
16:51:54 Quit tvelocity[a] (Read error: 110 (Connection timed out))
16:52:18 Join tvelocity[a] [0] (
16:53:44CIA-43New commit by zagor (r22593): Fixed checkwps build warnings. Updated checkwps makefile to be closer to main.
16:54:03 Join gregzx [0] (
16:54:09funmanI think only lowlight got a YH820
16:54:24 Quit Creposucre ("CGI:IRC")
16:55:19 Quit tvelocity[away] (Read error: 101 (Network is unreachable))
16:58:06 Join tvelocity[away] [0] (
16:58:56 Quit tvelocity (Read error: 101 (Network is unreachable))
16:59:12 Quit chandoo (Read error: 110 (Connection timed out))
17:00:18Zagoraw bugger, forgot to mention Rosso Maltese in the commit.
17:05:15 Quit Zagor ("Don't panic")
17:11:03 Join gregzx_ [0] (
17:11:43 Quit gregzx_ (Remote closed the connection)
17:13:44 Quit tvelocity[a] (Read error: 110 (Connection timed out))
17:20:19 Join cornwall [0] (
17:22:44cornwallhello! If there's anyone available and would like to help, I have a question :)
17:23:32TorneWuffdon't ask to ask, just ask
17:23:35 Nick TorneWuff is now known as Torne (
17:23:57cornwallalright, then, how do I customize what metadata is displayed on the Now Playing Screen?
17:24:12cornwallI can't seem to find any relevant files in .rockbox
17:25:21amiconnkkurbjun: I would expect the diretcion flip to be a command, not a pin. And how does the dma do blocks? That means it needs to skip pixels somehow
17:25:37GodEatercornwall: you would be needing to modify whatever .wps file your current theme is using
17:25:54 Quit gregzx (Connection timed out)
17:26:13cornwallaha, I see it
17:26:15amiconnAnd what does the width have to do with burst size? I would think it's the height, since you're going vertically
17:26:16cornwallthank you! :)
17:26:22*amiconn is slightly foncused
17:26:48GodEatercornwall: might prove useful
17:27:18cornwallGodEater, thank you very much. I appreciate it
17:35:07 Join NuNuRs [0] (
17:36:15NuNuRshello :)
17:38:02 Join tvelocity[a] [0] (
17:39:15 Nick dys` is now known as dys (n=andreas@
17:39:23cornwallGodEater, I'm trying to take the default .wps theme and just add a <composer> tag. It seems like this should work, shouldn't it?: %s%al%ic
17:41:58GodEaterah, I'm no expert on editing them
17:42:03GodEaterpixelma might be able to help
17:42:22cornwallGodEater, thanks anyway
17:45:21 Quit kugel (Read error: 110 (Connection timed out))
17:45:31 Join kugel [0] (n=kugel@rockbox/developer/kugel)
17:45:40pixelmacornwall: (a) which default WPS - cabbiev2 or the text based one (b) if it's cabbiev2 - which player do you have, which file do you edit and where do you put the line?
17:46:21pixelmaor answer the last questions anyways, might be some useful information in that
17:46:26 Quit pamaury ("exit(*(int *)0 / 0);")
17:46:28 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
17:47:12cornwallIt is cabbiev2 on an ipod 5.5G. I edited the .rockbox/wps/cabbiev2.wps file
17:47:15kkurbjunWamiconn, the dma controller multi-dimensioned - it is designed for image transformations with some of the co-processor interfaces so it can take in a stride along with an x/y offset and a width/height and properly do the transfers from memory.
17:47:17kkurbjunWThe LCD does not appear to have a controller that you speak to over an interface standard. The datasheet is here: I miswrote when I said width, you are correc that the height has an effect on the burst, I was thinking width in the portrait sense.
17:47:25cornwallI placed %s%al%ic right above the year tag
17:50:02amiconnkkurbjun: Hmm, and it can't use negative stride?
17:50:15 Quit Xerion (" ")
17:50:26kkurbjunWI would normally expect that the lcd inverstion would be with a command also, but the only thing that we have found are some commands that control the backlight. I've been able to change the resolution with a pin so some of what is defined in the datasheet is actually connected to the SOC.
17:51:04pixelmathere is a catch - cabbiev2 uses conditional viewports with this tag info, one is shown if album art is present, the other is shown without album art. Did you put your line into both parts (otherwise thee line looks fine)?
17:51:05kkurbjunWI don't think so, I can double check that
17:51:20pixelmathat was for cornwall
17:51:46cornwallboth parts? I must have missed that. I'm not using album art
17:51:49cornwalllet me check
17:53:38cornwallpixelma, great! it works well. Thank you very much
17:54:51pixelmayou're welcome :)
17:55:13 Join Xerion [0] (
17:55:14 Quit tvelocity[away] (Read error: 101 (Network is unreachable))
17:55:59 Quit cornwall ("Leaving")
17:57:25kkurbjunWamiconn, no it does not support negative values for the stride
17:57:28 Quit Zarggg ()
18:00:46 Quit kugel (Nick collision from services.)
18:00:54 Join kugel [0] (
18:03:06kkurbjunWamiconn, the bursts are in 16 word aligned (16 bit words) so the height of the LCD will work fine as a strip, if you are considering a fullscreen update, the processor will be updating the registers about 2x as often, but I think the majority of the time will still be spent waiting for the transfer to occur so it might not be a big performance hit.
18:05:59 Join tvelocity[away] [0] (
18:06:04 Quit funman ("free(random());")
18:07:17 Quit courtc (Read error: 113 (No route to host))
18:08:28 Quit kugel (Nick collision from services.)
18:08:34 Join kugel [0] (
18:11:57 Quit tvelocity[a] (Read error: 60 (Operation timed out))
18:12:29 Join tvelocity[a] [0] (
18:13:13 Join Zarggg [0] (
18:17:23pixelmahmm, if he keeps the name cabbiev2 it will probably be overwritten on his next Rockbox update... guess he'll figure it out
18:17:43***Saving seen data "./dancer.seen"
18:19:51 Join NuNuRs`tmp [0] (n=nunurs@
18:26:58 Quit kugel (Nick collision from services.)
18:27:04 Join kugel [0] (
18:27:10 Join bertrik [0] (
18:29:10 Quit tvelocity[away] (Read error: 110 (Connection timed out))
18:29:38 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:31:37 Join JdGordon_ [0] (i=ae915b9d@gateway/web/freenode/x-yfrbkzpeyohnwchi)
18:34:21 Quit kugel (Remote closed the connection)
18:34:47 Join funman [0] (n=fun@rockbox/developer/funman)
18:36:04 Quit Zambezi (Remote closed the connection)
18:38:26 Quit NuNuRs (Read error: 110 (Connection timed out))
18:38:27 Nick NuNuRs`tmp is now known as NuNuRs (n=nunurs@
18:40:38 Quit JdGordon_ (Ping timeout: 180 seconds)
18:41:35 Join domonoky1 [0] (
18:41:50 Join JdGordon_ [0] (i=ae915b9d@gateway/web/freenode/x-vsxplnsfxdtyabvi)
18:52:03 Join Zambezi [0] (i=Zulu@
18:59:00 Quit domonoky (Read error: 110 (Connection timed out))
19:10:10 Join PaulJam [0] (
19:10:54 Quit JdGordon_ (Ping timeout: 180 seconds)
19:13:08 Join Xerion_ [0] (
19:16:56 Join DarkDefender [0] (
19:17:33 Join merbanan [0] (
19:18:16 Join JdGordon| [0] (n=Miranda@nat/microsoft/session)
19:21:25 Quit kkurbjunW (Remote closed the connection)
19:21:39 Join kkurbjunW [0] (n=karlk@
19:29:05 Join Ubuntuxer [0] (
19:29:22 Quit Xerion (Connection timed out)
19:29:24 Nick Xerion_ is now known as Xerion (
19:33:47 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:33:56 Join Lear [0] (i=chatzill@rockbox/developer/lear)
19:36:25 Join tvelocity[away] [0] (
19:37:33 Quit tvelocity[a] (Read error: 60 (Operation timed out))
19:39:01 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
19:41:02 Quit merbanan (Read error: 110 (Connection timed out))
19:48:01 Join stripwax [0] (
19:48:05 Join tvelocity[a] [0] (
19:48:19 Quit stripwax (Client Quit)
19:49:32 Join stripwax [0] (
19:49:51 Quit stripwax (Client Quit)
19:51:04 Join stripwax [0] (
19:51:26 Quit stripwax (Client Quit)
19:55:20funmanI wonder if the yh920 led can be controlled or if it's always on
20:04:02 Quit tvelocity[away] (Read error: 110 (Connection timed out))
20:04:03 Join gregzx [0] (
20:05:44 Join Thundercloud [0] (
20:07:02 Quit Ubuntuxer ("Leaving.")
20:14:51 Quit tvelocity[a] (Read error: 110 (Connection timed out))
20:15:43rasherHow long's a long buttonpress?
20:17:46***Saving seen data "./dancer.seen"
20:18:06funman30 ticks (firmware/drivers/button.c: REPEAT_START)
20:18:43 Quit PaulJam (".")
20:19:57 Join chandoo [0] (
20:20:17rasherHm, I guess I don't have to worry about that as long as I'm inserting my own events
20:21:01rasherMan, I still can't make invadrox exit using this method
20:24:31 Join GreatBeaver [0] (
20:24:42GreatBeaverhow come my mp3 player has a delay every time i try to pause the music now?
20:25:02GreatBeaveri think i remember that it sometimes would stop instantly when i click pause
20:25:05JdGordon|turn off fade on pause
20:25:07GreatBeaverbut now it always has a 1 second delay
20:25:15GreatBeaverit isnt a fade, it stops completely
20:25:20GreatBeaveri'll try to disable that
20:25:35 Join CaptainKwel [0] (i=2669ecc2@gateway/web/freenode/x-wxtpblcbwnnnzzjn)
20:26:02GreatBeaveri wonder if maybe my mp3 player's battery is expanding and is applying presure on my harddrive or the zif cord
20:26:07GreatBeaverand cutt off some of the wires
20:33:54 Quit DarkDefender (Remote closed the connection)
20:34:57 Join TechGuy [0] (
20:36:25 Quit chandoo_ (Read error: 110 (Connection timed out))
20:36:30 Join chandoo_ [0] (
20:43:28 Quit pixelma (Nick collision from services.)
20:43:28 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
20:43:48 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
20:44:33 Quit amiconn (Nick collision from services.)
20:44:36 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
20:44:56 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
20:53:20 Quit chandoo (Read error: 110 (Connection timed out))
20:56:46 Quit Thundercloud (Remote closed the connection)
21:00:50 Quit Lss (Read error: 104 (Connection reset by peer))
21:01:47 Join merbanan [0] (
21:03:35 Join petur [50] (n=petur@rockbox/developer/petur)
21:06:34 Part GreatBeaver
21:11:00 Join Topy [0] (
21:13:16rasherWhat does button_status() do compared to button_get() ?
21:13:39funmanbutton_status gives the raw status while button_get is a higher level
21:13:56funmanlike you can use timeouts, repeats, etc
21:14:26funmanso button_status()|BUTTON_REPEAT is not gud
21:16:39rasherThis is not good for my button-faking adventures
21:25:44rasherCan I push something into the place where button_status() reads from?
21:25:51JdGordon|rasher: how about having the listener stick the incoming buttons into a queue and then setup a proper button_read_device() type thing?
21:25:59 Join HBK [0] (
21:26:11JdGordon|or.... make the sim a real target... RaaA style
21:26:15rasherJdGordon|: the sim doesn't even do that
21:26:34JdGordon|make it do that :)
21:29:01 Quit T44 (Read error: 113 (No route to host))
21:29:22rasherdon't wanna
21:33:33JdGordon|as sucky as it is.. thats the best way to have it work for real :)
21:34:26rasherJdGordon|: you are aware that the sim doesn't even use the real button.c?
21:34:48JdGordon|make the sim a real target so it does
21:34:55rasherTotally out of scope
21:35:05rasherI just need button_status() to return something sensible :)
21:35:07JdGordon|bring it into scope :)
21:37:28 Join courtc [0] (n=court@unaffiliated/courtc)
21:40:43 Join Hillshum [0] (
21:42:59 Quit KBH (Read error: 110 (Connection timed out))
21:44:11TechGuywhat all can be done iwth the Z80 emulator?
21:44:13TechGuyfor Rockbox
21:44:34gevaertsemulate a Z80?
21:44:42JdGordon|beaten :p
21:50:03 Quit Lear ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
21:51:06amiconnbutton_status() reads the current status from the hardware (or shadow variables), button_read() reads from the queue
21:52:36amiconnbutton_status() is available to plugins only for those emulator plugins (e.g. rockboy). Using it elsewhere is frowned upon
21:54:40amiconnrasher ^
21:56:19amiconnNotice that if a plugin uses both long and short press using button_get(), i.e. without the action api, you need to make sure you send the release event after a press and an optional repeat, otherwise the plugin won't react properly to the events
21:57:15amiconnThe release might be unnecessary for a long press if it's detected as the press-repeat sequence, i.e. without repeating the action
21:57:45rasherinvadrox uses button_status. Naughty!
21:58:20amiconnIt really shouldn't, being a non-emulator plugin
21:58:42rasherI managed to tell the sim which button was pressed, so it can respond properly to button_status() now
21:59:17 Quit LambdaCalculus37 ()
21:59:27rasher(just by setting the btn variable)
22:00:09CIA-43New commit by rob (r22594): D2: Move ACTION_WPS_STOP from the power button as it's too easy to hit by accident when disengaging the hold switch. Use a long press of the top-right ...
22:00:11 Join evilnick [0] (i=208bb833@gateway/web/freenode/x-cfwkynfchmvzhcly)
22:00:35rasherOther offenders: Brickmania, chopper, pong and rockblox1d (!)
22:00:43rasherAlso wav2wv
22:01:27rasherIt seemed that at least invadrox was simply doing the equivalent of testing for |BUTTON_REPEAT
22:02:15*amiconn wonders whether this might be necessary (or at least better) for some real-time cases
22:17:50***Saving seen data "./dancer.seen"
22:19:00rasherI'm not even qualified to speculate
22:19:21rasherwouldn't consider rockblox1d real-time critical though
22:19:59 Quit funman ("free(random());")
22:20:16rasherI think that's a misunderstanding. It's even comparing to "ONEDROCKBLOX_DOWN|BUTTON_REPEAT" and "SYS_USB_CONNECTED". I doubt button_status() woudl return those?
22:20:58 Quit gregzx ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
22:21:00rasherAnd indeed, it does not react to usb-insertion
22:21:26 Join kick_me [0] (n=57a0c3ef@gateway/web/cgi-irc/
22:25:14 Quit n1s ("Lmnar")
22:27:44 Quit evilnick (Ping timeout: 180 seconds)
22:28:14 Quit GeekShadow ("The cake is a lie !")
22:28:53 Quit merbanan (Read error: 110 (Connection timed out))
22:30:06 Join Thundercloud [0] (
22:33:57 Join froggyman [0] (
22:42:07 Join Llor_Phone [0] (n=Llorean@
22:43:26 Join Ubuntuxer [0] (
22:44:03 Quit Ubuntuxer (Client Quit)
22:49:26AlexPUm, what is the first one on this page?
22:49:29 Join Zagor [242] (
22:50:01gevaertssomeone trying to get round the rules?
22:50:18AlexPYeah, we require names as well don't we?
22:50:37gevaertsthat too
22:51:45AlexPI think it should go then
22:52:05Llor_PhoneCopyright image, anon submitter. Seems clear.
22:52:33Llor_PhoneAs a note, the theme site seems to work well on my phone
22:52:46froggymanis that theme intended for use with the OF then?
22:53:54HillshumDo you just take it down?
22:54:09Llor_PhoneWhatever it is, it should be removed and possibly the verification tightened.
22:54:33AlexPIt's gone
22:54:42AlexPHillshum: Yes, I just deleted it
22:54:54Llor_PhoneDid it have a wps?
22:55:05Llor_PhoneOr, how did it get in?
22:55:38AlexPrasher: As a note, should/could the theme site have a works with 3.3 tag, as well/instead of 3.2?
22:55:52Llor_Phoneinstead of
22:55:53rasherAlexP: we don't hav e a checkwps for 3.3
22:56:12Llor_PhoneCan we make one?
22:56:17rasherSo we're stuck waiting until someone backports the fix or 3.4 is released, whichever comes first
22:56:21rasherLlor_Phone: it didn't build at the time
22:56:44Llor_Phone3.4 is so soon anyway
22:56:56Llor_PhoneBy tis point it's not wort it.
22:59:06JdGordon|are all themes rescanned when a new checkwps is done?
22:59:10 Quit Llor_Phone (Remote closed the connection)
22:59:29JdGordon|or at least for each new release?
22:59:51JdGordon|I dont tinhk there is any point keeping old themes are if they dont work... (it would be good if they were avilable so people can fix them though)
23:03:01rasherAll themes are rescanned daily
23:04:17 Quit petur ("Zzzzzz")
23:04:45 Join gevaerts-mobile [0] (n=c3905c89@gateway/web/cgi-irc/
23:04:47rasherIt takes a few minutes
23:05:26 Quit gevaerts-mobile (Client Quit)
23:06:38JdGordon|what happens if they stop working in the current build?
23:13:27 Nick fxb is now known as fxb__ (
23:15:46pixelmaJdGordon|: you'll find a notice in red saying that this theme doesn't work with the current build, you can see some on the c200 page and I guess others have it as well
23:17:55JdGordon|ok, cool
23:18:22*JdGordon| annoys someone about getting the themes skin usage stats on the webpage again :)
23:19:35JdGordon|unless of course thats not something that people tihnk users will find useful?
23:21:16 Quit kkurbjunW (Remote closed the connection)
23:21:41HillshumJdGordon|: the goal isn't to be useful to the users!
23:21:42 Join kkurbjunW [0] (n=karlk@
23:21:52pixelmawhich reminds me of that weird result with a half broken WPS - the SansaPod theme on the c200 page has 2 errors, one is an additional e infront of the %X tag and the other is an %xd without %xl. With both errors I got the complete default WPS, fixing the e%X gave me something that looked like a blanked out default WPS with only the progressbar left
23:22:24rasherJdGordon|: I don't see why users would find it useful
23:23:52JdGordon|pixelma: we could probaly put error checking in for the missing %xl... dunno why e%X failed though
23:23:57 Join mcuelenaere [0] (
23:24:08JdGordon|rasher: at some point in the future its very possible users will need that info
23:24:13rasherJdGordon|: Why?
23:24:37JdGordon|pixelma: if you have a chnace, can you have a look at the sub/lines patch of mine in the tracker please? see if any themes break?
23:24:45pixelmaI am fine with it failing, just wondered why showing the default WPS was broken then
23:29:13 Quit Rob2223 (Read error: 104 (Connection reset by peer))
23:29:53 Join Rob2222 [0] (
23:34:50 Join Llorean [0] (
23:35:48CIA-43New commit by rob (r22595): D2: Enable ARM cache coherency functions (eg. during codec load), which should eliminate data aborts/freezes on track changes. NOTE: The linker script ...
23:36:42mcuelenaerewhat exactly triggers the empty svn table on the front page upon new commits?
23:37:27 Quit bmbl ("Bye!")
23:38:23pixelmaI've seen it sometimes empty too but an immediate reload helped, didn't see a connection to commits yet but maybe there is
23:39:12pixelmaI also saw the same with the Mail table a few times by the way
23:39:40mcuelenaereimmediate reload on Chrome (Linux) didn't seem to help (after ~10 secs it went away)
23:39:56rasherIt's probably just rebuilding the table
23:40:08mcuelenaereprobably, but couldn't it flush its output after it's done?
23:43:29 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
23:44:08 Quit jgarvey ("Leaving")
23:45:27 Quit bertrik ("De groeten")
23:47:38 Quit flydutch ("/* empty */")
23:49:59 Join andrewrb [0] (
23:57:47CIA-43New commit by rob (r22596): D2: Obey the SD card write-protect tab

Previous day | Next day