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 2010-08-17

00:05:10 Quit krazykit (Ping timeout: 276 seconds)
00:15:19 Quit petur (Quit: Leaving)
00:15:39 Quit bertrik (Quit: :tiuQ)
00:15:39S_a_i_n_tis broken for anyone else?
00:15:54S_a_i_n_tThe page displays, but, no logs. :/
00:16:12Jennifurit works
00:16:41S_a_i_n_tHmmmm...all the other pages work for me in Chrome except for todays.
00:16:53S_a_i_n_tI'll try IE8
00:17:33Jennifuri'm using opera
00:21:04S_a_i_n_tHmmmm, todays log works for me in IE8 but not Chrome.
00:22:38 Quit AndyIL (Ping timeout: 248 seconds)
00:24:39S_a_i_n_tmamarley; The answer to you "why do animations run quicker when there is user input?" is because a buttonpress triggers a fullscreen update.
00:25:02S_a_i_n_tSo you're forcing the screen to update more frequently than it normally would.
00:25:30mamarleyThanks. The thing I was mainly wondering about was the disk activity indicator on your theme, I noticed that it ran much more slowly when I wasn't pressing a button.
00:26:41S_a_i_n_tyes, it alternates at 0.1 second due to a buttonpress triggering a fullscreen update scrolling through a list will make it alternate *much* more smoothly.
00:27:05 Quit CGL (Ping timeout: 240 seconds)
00:27:59S_a_i_n_tIs there anything you think I should change/implement on that theme? I'll take any critique, good or "bad".
00:28:42 Join AndyI [0] (~pasha_int@
00:29:06mamarleyThe only thing I noticed was a conflict between the disk activity indicator and the play/pause indicator. If I remember correctly, the OF only displays the disk activity indicator when the iPod is plugged in, but I might be remembering wrong.
00:29:51S_a_i_n_tYes, you're correct...but I thought it needed a disk access icon in all screen instead of just in the USB screen.
00:30:05 Quit bluebrother (Ping timeout: 240 seconds)
00:30:08 Join bluebrother [0] (
00:30:08 Quit bluebrother (Changing host)
00:30:08 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
00:30:15S_a_i_n_tI put in where it is, because that is its position in the USB screen.
00:30:52amiconnOh my... more jpg bugs
00:30:58S_a_i_n_tOn a flash target, it never runs for too long...I find it quite cool actually.
00:31:31S_a_i_n_tOn the iPod Video theme I am making I might add an option to disable it...but I'm not sure yet.
00:32:14mamarleyI like it very much though.
00:32:21S_a_i_n_tmamarley: Is there anything else?
00:32:49S_a_i_n_tI'm rather keen to see how it looks with aliased fonts...if you could take a screenshot I would appreciate it.
00:33:04mamarleyNo, not until there is support for fancy sliding transitions. ;)
00:33:27S_a_i_n_tHehe, as soon as I can do it...I will ;)
00:33:35mamarleySure (about the screenshot.) How do I to that?
00:35:41S_a_i_n_tgo to System - Debug - Screendump and press Select. you'll get a splash saying "Screendump Enabled", then when you plug in to USB it will capture a screenshot and save it to root instead of connecting via USB.
00:36:11S_a_i_n_tThen, when you want to connect via USB, go to the screendump setting again and "Select" it again to disable it.
00:36:35S_a_i_n_t(you'll get another splash saying it is disabled)
00:36:51mamarleyGreat! When I did that, I got "*PANIC* Stkov usb"
00:37:18S_a_i_n_tAh....that's right.
00:37:25S_a_i_n_tI forgot about that bug...sorry.
00:38:09*mamarley 's iPod was crashed by S_a_i_n_t!
00:38:11S_a_i_n_tDo you find you often get *PANIC* Stkov NAND?
00:38:27mamarleyAnd it is fine, I have iLoader so I don't get the WSoD.
00:38:32S_a_i_n_tHmmmm, what capacity is your Nano?
00:38:50mamarleyWait, I guess the Panic screen is a WSoD. My nano is 4gb.
00:39:33TheSevenmamarley: WSoD = usually the "Use iTunes to restore" crap
00:39:58mamarleyI know.
00:40:23TheSevenS_a_i_n_t: I've never seen a stkov nand either
00:40:28mamarleyBut the panic screen technically is a WSoD. By the way, I love iLoader. It is extremely fast and error-less.
00:40:38mamarleyThanks for writing it.
00:41:25TheSeventhere will be a new release soon :)
00:41:34mamarleyWhat new features will there be?
00:41:49mamarleyOr is it just a bugfix release?
00:41:52TheSevenit will mostly be featuring easier (un)installation, and better file i/o handling for themes
00:42:05TheSevenbut it will be a tiny bit slower
00:42:41TheSevenintegrating emBIOS will be great for developers though :)
00:43:01*S_a_i_n_t wonders what emBIOS is...
00:43:48 Quit Jerom (Quit: Leaving.)
00:45:03TheSevenan os kernel, hardware abstraction layer, standard library, and debugger
00:45:23TheSeventhis will be replacing ibugger
00:45:54mamarleyTheSeven: Any news on the nano2g vsync interrupt?
00:46:28*mamarley hates busywaiting :)
00:46:34 Quit bluebrother (Read error: Connection reset by peer)
00:46:48 Join bluebrother [0] (
00:46:48 Quit bluebrother (Changing host)
00:46:48 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
00:47:33S_a_i_n_tmamarley: The reason I asked about *PANIC* Stkov NANDs is because I had to change the nand stack size to [0x1000] to not get it to *PANIC* whilst doing simple things like browsing the menu, or loading a theme.
00:47:36***Saving seen data "./dancer.seen"
00:47:57S_a_i_n_tIt may have just randomly fixed itself though, perhaps I don't need it that high anymore.
00:48:18mamarleyI haven't had any Stkov NANDs except for awhile back when the EABI compiler caused them. I think you fixed that though.
00:48:27mamarleyWhat does Stkov mean?
00:48:30TheSevenstack overflow
00:48:37S_a_i_n_t"Stack Overflow"
00:48:50amiconnStkov usb on screendump? I haven't seen that on any target so far
00:49:28TheSeven2048 sector size + on-stack sector buffers in the FAT code :/
00:49:46S_a_i_n_tamiconn: On my Nano2Gs it just crashes to a white screen...without printing an error.
00:49:55 Quit jfc (Ping timeout: 240 seconds)
00:50:01mamarleyCould that be worked around by making the stack bigger?
00:50:03S_a_i_n_tI have not tried it recently, however.
00:50:11amiconnI know. An intermediate fix would be to increase the problematic stacks
00:50:42S_a_i_n_tmamarley: That is exactly what I have done...more than doubled the nand stack, but it is a workaround...not a fix.
00:50:49n1samiconn: that's one of those "temporary" fixes that tend to stick around ;/
00:51:06S_a_i_n_tAh, you meant the USB stack.
00:52:03S_a_i_n_tWhere is the USB stack size defined for the Nano2G?
00:52:10amiconnn1s: I'm not a fan of temporary workarounds, but sometimes there is no real choice
00:52:54mamarleyMe neither.
00:52:58 Quit Judas_PhD (Quit: Aaarrrggghhh!!!)
00:52:59S_a_i_n_tI'd be happy to test to see how large it needs to be without *PANIC*ing
00:53:18S_a_i_n_tI'm not sure where the nand stack is defined though.
00:57:07 Join Judas_PhD [0] (
00:57:26 Quit Judas_PhD (Client Quit)
00:58:20mamarleyI don't know where it is defined either. I am not even sure I would recognize it if I saw it.
00:58:35 Quit bmbl (Quit: Bye!)
00:59:05 Quit S_a_i_n_t (Read error: No route to host)
00:59:39 Join krazykit [0] (~kkit@
01:03:47 Join S_a_i_n_t [0] (S_a_i_n_t@
01:08:51 Part toffe82
01:09:44 Join designate72 [0] (
01:11:43TheSevenusbcore.c IIRC
01:14:54 Quit JdGordon (Ping timeout: 248 seconds)
01:23:51 Quit anewuser (Quit: FAMICOMPO 7 - August 15 Through August 31 JST -
01:26:36 Join jfc [0] (
01:31:03S_a_i_n_tTheSeven: Thankyou.
01:32:43 Quit yalu (Ping timeout: 240 seconds)
01:34:44 Join yalu [0] (
01:35:04 Quit Dreamxtreme (Quit: Don't follow me)
01:35:41 Quit DerPapst (Quit: Leaving.)
01:38:49 Join Strife89 [0] (
01:39:40 Quit jordan` (Quit: Coyote finally caught me)
01:44:57 Quit Ludo6431 (Quit: EOF)
01:46:50 Join jordan` [0] (
01:56:51 Quit n1s (Quit: Lämnar)
02:03:07 Join CGL [0] (~CGL@
02:03:28 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
02:08:35 Quit GeekShadow (Quit: The cake is a lie !)
02:19:12 Join kugel [0] (
02:19:14 Quit kugel (Changing host)
02:19:14 Join kugel [0] (~kugel@rockbox/developer/kugel)
02:24:04kugelmamarley: I had a look at chopper too once
02:24:21kugelIIRC I noticed you cannot make it more smooth without making it faster as well
02:24:39kugelbecause the game speed starts very low (at 3 or so)
02:24:53mamarleyIn my patch it starts at 2.
02:25:00mamarleyAny lower, and it is just too slow/easy.
02:25:29kugelI went for fixed point math approach, but I couldn't yet manage to keep the game balance entirely the same
02:26:22mamarleyI'm not sure what you mean... Did you also write a patch?
02:26:34kugeluhm, yes, but I've never posted it anywhere
02:26:50mamarleyOh, that's why I didn't know about it.
02:27:05mamarleyI am not sure how to used fixed point math, though.
02:27:06kugeldoesn't it increase the X speed by 1 after a few ticks? if yes, then you basically can't make it smoother without altering the balance and without fixed point math
02:27:44mamarleyIt does increase X speed.
02:28:30mamarleyAll I did was remove the CYCLETIME and adjust some other things to make it as close as I could to the original.
02:28:43kugelah right, here: if (chopCounter == 100){ iPlayerSpeedX++;
02:29:18kugelyou can't decrease chopCounter because X++ is the minimum possible increase (without fp)
02:29:35mamarleyYeah. I changed that to 1000 or 600 (depending on screen size) to compensate for the greater speed.
02:30:20kugelwell, I've maybe overlooked something, one reason why I never posted the patch :) (I also stopped playing chopper :P)
02:32:38 Quit mamarley (Remote host closed the connection)
02:32:58 Join mamarley [0] (
02:33:51mamarleyWhat was that reason?
02:33:54 Quit Jennifur (Ping timeout: 245 seconds)
02:35:30 Quit S_a_i_n_t (Ping timeout: 264 seconds)
02:35:34 Join [Saint] [0] (S_a_i_n_t@
02:36:26[Saint]mamarley: If you decreased the starting speed of your patch, it would be a lot better IMO
02:36:32kugelI couldn't get it to work without altering the game balance
02:36:40[Saint]then, I'd say it would have a chance of being committed.
02:37:13 Nick [Saint] is now known as S_a_i_n_t (S_a_i_n_t@
02:37:16mamarleyI would like to decrease the starting speed some, but decreasing it to 1 (from its current 2) makes it way too slow.
02:38:12mamarleyI guess I could remove the multiplier that is used for normal mode and set another distinct value. Let me try that.
02:38:49S_a_i_n_tmamarley: decreasing it from 2 to one should be sufficient.
02:39:14mamarleyThe multiplier or the starting speed?
02:39:42S_a_i_n_tstarting speed
02:40:14mamarleyI personally think that is too slow. It is slower than the start speed for an unpatched version.
02:40:28S_a_i_n_tOh...does it take a decimal value?
02:40:33S_a_i_n_tie. 1.5?
02:40:38mamarleyNope. Ints only.
02:40:46mamarleyBelieve me, I wish ARM had a FPU.
02:45:24TheSeventhere are some that have
02:45:43TheSevenbut afaik none of the rockbox targets (except maybe raaa)
02:47:37***Saving seen data "./dancer.seen"
02:48:44 Join Judas_PhD [0] (
02:48:55 Join bunnyboi [0] (
02:49:06JdGordon|doesnt the beast?
02:51:43Tornethe beast has vfpv2, yes
02:52:10Torneisn't the nano2g an arm11?
02:52:24mamarleyNo, it is ARM940T.
02:52:50mamarleyThe nano3g is ARM11, I think. Not sure, though.
02:53:08 Join Dreamxtreme [0] (Dreamxtrem@
02:53:30kugelarm11 doesn't necessarily mean fpi, does it?
02:53:35S_a_i_n_twith the keymaps...what is the third value in the tables for? ie. { ACTION_STD_FOO, BUTTON_FOO|BUTTON_BAR, BUTTON_BORK },
02:53:39S_a_i_n_twhat is "BUTTON_BORK"?
02:54:04JdGordon|not a real button :p
02:54:11kugeli.e. what buttons have to be pressed one tick before
02:54:14Tornekugel: technically no, the vfp11 cell is optional
02:54:26Tornekugel: but in practise every arm11 based SoC or device i've ever seen has it
02:54:56kugelthat's my experience too
02:54:57S_a_i_n_tI managed to swap quicksceen and the context menu for iPod *once*, but I can't seem to figure it out again.
02:55:06S_a_i_n_tI did it, then accidentally reverted it.
02:55:14S_a_i_n_tAnd I don't remember what I did :(
02:55:19Tornekugel: ARM allow for a lot of configurations that nobody even makes, let alone one that nobody uses
02:55:42Tornekugel: there's loads of permutations of ARMv7 floating point that just don't even exist, you can't configure ARM's macrocells to emit them ;)
02:55:53Tornesomeone could make them in theory but they won't
02:56:12kugelS_a_i_n_t: in your example, to execute ACTION_STD_FOO, you need at tick X BUTTON_FOO and BUTTON_BAR to be pressed at the same time, but at tick X-1 BUTTON_BORK (pressumably impossible to do)
02:56:49mamarleyOK, I just updated my chopper patch to slightly decrease the starting speed in normal mode. It used to be 2*2=4, now it is 3.
02:57:29S_a_i_n_tkugel: { ACTION_STD_FOO, BUTTON_FOO|BUTTON_REPEAT, BUTTON_BORK }, is a better example...signifying that FOO needs to be held.
02:57:55S_a_i_n_t(that was what BAR was supposed to be, but I forget that would mean "press both at once"
02:58:47kugelBUTTON_FOO and BUTTON_BORK are still different buttons, it still seems impossible :)
03:00:01S_a_i_n_treal example: (edited to suit) "{ ACTION_STD_CONTEXT, BUTTON_MENU|BUTTON_REPEAT, BUTTON_MENU }"
03:00:22S_a_i_n_tI was wondering what the last value was, "BUTTON_MENU"
03:00:26kugeland that doesn't work?
03:00:34S_a_i_n_tDon't know yet.
03:00:51S_a_i_n_tI messed up the end value...that's why I was wondering what it was.
03:00:59kugelwell, after some ticks BUTTON_REPEAT is added to the current button (I don't know the exact number) if you held it for a while
03:01:07S_a_i_n_tI forget to swap the last value around, as well as the buttons.
03:01:39kugelBUTTON_MENU as precondition means that one tick before BUTTON_REPEAT was added BUTTON_MENU was the only button (which is true for 99.999999% of the time)
03:02:02 Quit komputes (Remote host closed the connection)
03:02:20S_a_i_n_tI need to recompile it other news, my efforts to add more colours to "lamp.rock" was entirely successful. \o/
03:02:50S_a_i_n_tI now have white, red, orange, yellow, green, blue, indigo, violet, and black.
03:02:54kugelthe pre-condition makes sense where you have different actions on a short and long press
03:03:13S_a_i_n_tAh, right...I get it now....*kinda*
03:03:28S_a_i_n_tIf my changes don't work this time, I shall harrass an expert.
03:03:30kugele.g. BUTTON_X|BUTTON_REL will only a trigger a short press, because in a long press the button value would be BUTTON_X|BUTTON_REPEAT|BUTTON_REL
03:04:09kugellikewise, BUTTON_REPEAT in the pre-condition will never trigger on a short press
03:04:31 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
03:05:00S_a_i_n_tAll I'm trying to do is make "Menu == Context Menu" and "Select == Quickscreen"
03:06:31S_a_i_n_tyes, that is what I have...*now*
03:06:39S_a_i_n_tI missed the precondition last time.
03:07:31kugelIIUC the pre-condition shouldn't matter in these cases (if they're BUTTON_NONE, if they're a real button they might indeed mess up things)
03:08:55 Quit piotrekm (Quit: piotrekm)
03:10:17 Quit perfectdrug (Ping timeout: 252 seconds)
03:13:51 Quit evilnick- (Read error: Connection reset by peer)
03:17:27 Quit designate72 (Ping timeout: 260 seconds)
03:20:49 Quit fyrestorm (Read error: Connection reset by peer)
03:29:30mamarleyS_a_i_n_t: Can you try my updated chopper patch?
03:30:26S_a_i_n_tmamarley: Not right this second, but, yes.
03:31:34mamarleyI am also considering setting the default start speed lower conditionally for the Gigabeats. I would also like it if somebody could try it on some of the Archoses.
03:35:44 Quit storm` (Ping timeout: 240 seconds)
03:35:47 Join storm` [0] (
03:36:54 Quit angelwolf71885 (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
03:42:46S_a_i_n_tkugel: Do you have a revised list of the icons you want me to change, and, have you been using the iconset on device yet?
03:43:03kugelnot yte
03:43:24S_a_i_n_tAh, try it out...*then* tell me if anything needs changing.
03:45:50JdGordon|anyway, a touch cabbie sbs would be very useful
03:48:34 Join CaptainKwel [0] (
03:50:36JdGordon|kugel: also, can we set a root dir for the browser?
03:57:37kugelthere was a short discussion, and it was disliked
03:58:05kugelbut the consensus seemed to be to offer some user-settable shortcuts (e.g. <Music>) or so
03:59:51 Join BRi7X [0] (
04:00:39JdGordon|when was this discussion?
04:01:11BRi7Xhowdy, does anyone know how to remove a ZIF ribbon from a zif hard disk? it feels like it won't come out when i tug on it... i'm afraid of ripping the ribbon, should i just go for it or what?
04:02:03 Quit mamarley (Remote host closed the connection)
04:04:08S_a_i_n_tBRi7X: Sometimes, there is a small lever, or catch...other times there is not, and it is just in there very snugly
04:04:21JdGordon|kugel: kugel iconset: /data/data/org.rockbox/rockbox/icons/icons.bmp should work no?
04:04:26S_a_i_n_tIf there is no obvious catch, just give it some force, but, not too much.
04:05:04 Quit sasquatch (Ping timeout: 276 seconds)
04:05:40kugelJdGordon|: /sdcard/rockbox/icons/foo.bmp should work as well, but I think it changes it back to the other path when saving settings (but that doesn't matter because it looks in /sdcard/rockbox first)
04:06:14kugelthe settings code doesn't work with two directories for the filename settings (because it only saves the basename without path and extension in memory)
04:06:51mc2739JdGordon|: in settings->theme settings->radio screen the current selected fms is not highlighted like wps and base. do you know what is needed to fix that?
04:07:21JdGordon|ask rasher, IIRC he made that work
04:07:23 Join kugel_ [0] (
04:07:45 Quit kugel (Disconnected by services)
04:07:49 Nick kugel_ is now known as kugel (
04:07:53 Quit kugel (Changing host)
04:07:53 Join kugel [0] (~kugel@rockbox/developer/kugel)
04:08:04JdGordon|does the yes/no screen not work at all in stylus mode?
04:10:31 Join literal [0] (
04:13:33JdGordon| victory is mine!
04:13:53JdGordon|manually copied S_a_i_n_t's icons into the data dir as tango_small.bmp :)
04:14:23JdGordon|S_a_i_n_t: yeah, they look good
04:15:06JdGordon|the "settings" menu has a - icon instead of the spanner/screwdriver thoguh
04:15:40kugelJdGordon|: it really should work under /sdcard/rockbox/icons, if not it's a bug
04:15:45 Join anewuser [0] (anewuser@unaffiliated/anewuser)
04:15:52JdGordon|well sure
04:17:34BRi7XS_a_i_n_t, yeah i think i got it... but any way i have it inserted the Gigabeat S won't power the hard drive up
04:18:52 Quit MethoS- (Read error: Connection reset by peer)
04:19:58 Join sasquatch [0] (
04:21:29BRi7Xif anyone's willing to trade me a ZIF hard drive for two 40GB IDE hdd's (for gigabeats). please, don't hesitate to message me. i'm desperate at this point! hah... ebay screwed me again (again).
04:34:31*kugel doubts r27834 works, but is happy to see contributions :)
04:41:04 Quit TheSeven (Ping timeout: 246 seconds)
04:41:32 Join Daniel [0] (
04:44:23 Part Daniel
04:45:05 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:45:06 Join Nerdy_ [0] (
04:45:27 Nick Nerdy_ is now known as Nerdy3_14159265 (
04:45:48Nerdy3_14159265Hey everyone
04:46:47Nerdy3_14159265I had a question about rockbox
04:47:40***Saving seen data "./dancer.seen"
04:47:53krazykitbut not anymore?
04:48:00*JdGordon| was too slow :)
04:48:24Nerdy3_14159265I'd get you if I didn't have to ask you guys something
04:48:48krazykitwell, it's much simpler to just ask rather than state the obvious :)
04:49:36 Join Barahir [0] (
04:49:52Nerdy3_14159265I was wondering if there was a way to play an album in rockbox and then have it automatically play another from the same artist (or different if set) in the correct order
04:50:25JdGordon|of course
04:51:16Nerdy3_14159265how then?
04:53:04 Quit kugel (Remote host closed the connection)
04:53:25 Quit Barahir_ (Ping timeout: 264 seconds)
04:53:40 Quit amiconn (Disconnected by services)
04:53:42 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:54:02 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:54:28 Quit pixelma (Disconnected by services)
04:54:30 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:54:41Nerdy3_14159265ummm, that doesn't seem to describe rockbox automatically selecting an album at random to play
04:54:46 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:57:05krazykitsounds like you want random folder advance
05:00:33Nerdy3_14159265I tried that but that didn't work like I'd hope, my songs aren't in the correct order when playing
05:04:48 Join fyrestorm [0] (
05:23:07 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
05:27:53 Quit fyrestorm (Read error: Connection reset by peer)
05:39:40 Quit shai (Ping timeout: 240 seconds)
05:42:59 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:48:27S_a_i_n_tJdGordon|: The "-" made sense to me more than the "spanner & screwdriver" as the "-" correlates to "submenu entered" and traditionally, in SVN cabbie icons, the "submenu" icon is a "+"
05:49:10S_a_i_n_tso, I figured, if submenu == +, then an icon representing the fact that you can go back one layer being a "-" made sense.
05:49:12JdGordon|yes, but the tools one is more obviously settings.. i tihnk
05:49:27S_a_i_n_tJdGordon|: but, it's not just for settings.
05:49:33JdGordon|I'm talking about "settings" in the top level menu
05:49:33S_a_i_n_tit's for all submenus
05:50:02S_a_i_n_tthe + is "submenu", the "-" is "submenu entered".
05:50:05JdGordon|wouldnt it be nice if the icons wernt so random :)
05:50:12S_a_i_n_tI can show you the icon layout if you want.
05:50:19S_a_i_n_tie. what icon is associated to what.
05:51:19JdGordon|na, irs fine, jjust thought the tools one would be better there
05:51:27S_a_i_n_tIf I change it to reflect the "settings" icon being a spanner & screwdriver, then *all* submenus would have that icon...which would make less sense IMO
05:51:28JdGordon|otherwise it loks good
05:51:46S_a_i_n_tI understand what you're getting at though.
05:52:03S_a_i_n_tIt's just that fixing one instance would "break" many others
05:52:15S_a_i_n_t(I *did* actually think about this ;))
05:52:29JdGordon|suuure you did :D
05:55:50 Quit cjcopi (Ping timeout: 264 seconds)
06:00:18 Join shai [0] (
06:07:21 Join cjcopi [0] (
06:11:40 Quit Nerdy3_14159265 (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
06:14:34 Quit kkurbjun (Changing host)
06:14:34 Join kkurbjun [0] (~kkurbjun@rockbox/developer/kkurbjun)
06:18:16 Quit BRi7X (Read error: Connection reset by peer)
06:18:44 Join BRi7X [0] (
06:25:47 Quit BRi7X (Read error: Connection reset by peer)
06:26:20 Join BRi7X [0] (
06:33:37 Join h3po [0] (
06:33:47 Quit anewuser (Quit: FAMICOMPO 7 - August 15 Through August 31 JST -
06:36:55 Quit Horscht (Quit: Verlassend)
06:47:42***Saving seen data "./dancer.seen"
06:55:38 Quit h3po (Quit: Leaving.)
06:55:55 Quit bunnyboi (Ping timeout: 276 seconds)
07:06:35 Join bunnyboi [0] (
07:16:46froggymanwhy is the Gigabeat S listed as a functional port on the main wiki page, but unstable on the main page?
07:22:57JdGordon|different requirements?
07:31:42 Quit t0rc (Quit: Give someone code, help them with one project. Teach someone to code, help them rule the world.)
07:40:00 Quit bunnyboi (Quit: Never look down on someone unless you're helping them up.)
07:42:01 Join stoffel [0] (
07:44:51 Join einhirn [0] (
08:03:40 Quit CaptainKwel (Quit: Ex-Chat)
08:08:02 Join advcomp2019_ [0] (~advcomp20@unaffiliated/advcomp2019)
08:10:00 Quit stoffel (Remote host closed the connection)
08:11:26 Quit advcomp2019 (Ping timeout: 265 seconds)
08:11:26 Quit z35 (Ping timeout: 265 seconds)
08:11:26 Quit avacore^ (Ping timeout: 265 seconds)
08:12:14 Join avacore [0] (
08:12:55 Join CIA-88 [0] (~CIA@
08:13:52 Quit cjcopi (Ping timeout: 264 seconds)
08:13:53 Quit CIA-9 (Ping timeout: 264 seconds)
08:13:53 Join cjcopi [0] (
08:14:28 Join z35 [0] (
08:26:34 Join LinusN [0] (linus@rockbox/developer/LinusN)
08:39:42 Join bieber_ [0] (
08:39:42 Quit bieber (Read error: Connection reset by peer)
08:40:12 Nick bieber_ is now known as bieber (
08:44:53 Quit bieber (Remote host closed the connection)
08:46:21 Join bieber [0] (
08:47:46***Saving seen data "./dancer.seen"
08:52:13S_a_i_n_tfroggyman: "functional" could really mean anything, it's quite possible that it could be functional, *and* unstable.
08:52:57S_a_i_n_tfunctional probably just means "the parts that work, function as intended"
08:54:14S_a_i_n_t"Unstable" doesn't necessarily mean "not stable" but (by my perception of the main page) "some things may be missing"
09:03:07 Join ender` [0] (
09:03:29S_a_i_n_tamiconn: ping?
09:04:21S_a_i_n_tDid you check if commenting/ifdef'ing out the codeblock I pointed out in pictureflow.c makes it run beter on your player?
09:04:40S_a_i_n_t(lines 1808, 1809, and 1810)
09:05:48S_a_i_n_tI ifdef'd !NANO/NANO_2G that block, and it runs fine again.
09:11:25S_a_i_n_tApparently that section of code really improves visual quality on some targets (according to the tracker entry about it), so I can't really see another way of doing it than a big mess of ifdef's for the targets known to have problems with it. (Nano 1G and 2G, the target you mentioned...possibly/probably others)
09:12:34S_a_i_n_tIf anyone can confirm whether pictureflow is running at radically reduced quality during playback, and on which would be helpful.
09:13:29 Join angelwolf71885 [0] (
09:17:15 Join petur [0] (d408b802@rockbox/developer/petur)
09:17:49 Quit mc2739 (Read error: Operation timed out)
09:18:05 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
09:28:19 Quit BRi7X (Read error: Connection reset by peer)
09:28:51 Join BRi7X [0] (
09:31:54 Join Rob2222 [0] (
09:35:32 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
09:35:50 Quit Rob2223 (Ping timeout: 276 seconds)
09:36:18 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:49:34 Join Jaykay_ [0] (
10:01:12 Nick Jaykay_ is now known as Jaykay (
10:22:49 Quit GeekShadow (Ping timeout: 240 seconds)
10:23:36 Quit JdGordon| (Quit: leaving)
10:24:27 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
10:26:38 Join JdGord [0] (
10:27:12 Quit GodEater (Quit: leaving)
10:27:18 Join GodEater [0] (
10:27:18 Quit GodEater (Changing host)
10:27:18 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
10:34:08 Join DerPapst [0] (~Alexander@
10:34:56 Join n1s [0] (~n1s@rockbox/developer/n1s)
10:40:44 Quit liar (Ping timeout: 260 seconds)
10:42:33 Join liar [0] (
10:45:27*amiconn wonders whether S_a_i_n_t has a different pictureflow.c than he has
10:46:12 Quit JdGord (Quit: Bye)
10:47:49***Saving seen data "./dancer.seen"
11:06:50 Quit delt_zZzZ (Ping timeout: 276 seconds)
11:14:00 Join watto [0] (~watto@
11:14:48 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
11:19:59JdGordonhere's a fun question...
11:20:37JdGordonwould anyone have problems with replacing the c code for some screens (where it makes sense obviously) with hardcoded skin code? (so no disk access)
11:20:54JdGordonBassically I'm talking about the time/date menu/screen thing which is entirely wasted code being in c
11:21:27 Join Ludo6431 [0] (
11:24:00n1sJdGordon: will voce work?
11:24:19JdGordonvoice doesnt work in that screen anyway
11:24:49n1soh, well could it work with the skin stuff then?
11:25:35JdGordonof course, all it does it display the clock, but I could do the same in about 3 lines of 3 and the const char* to load
11:25:48JdGordonI think :p
11:26:53n1sare you talking about set_time_screen r am i way off?
11:27:11n1ss/ r/ or/
11:28:25JdGordonway off
11:28:37n1sok, where are you then? :)
11:28:47JdGordonthe system > time & date
11:29:00JdGordonthe top half
11:29:39n1sah, so keep the looks and functionality but do the clock display with the skin engine?
11:31:03JdGordonyeah, and make it look better
11:31:18n1si can't think why anyone would be against that if it simplifies the code and or makes it smaller
11:31:49JdGordonim sure someone will have an argument against :)
11:32:54n1salso, i would have expected that screen to announce the current time but this change will not prevent anyone from adding that, right?
11:49:22 Join u42p [0] (
11:49:28u42phi, does someone have a use for a rio karma with broken volume up button and half-broken wheel?
11:51:41 Quit sasquatch (Ping timeout: 276 seconds)
11:59:43JdGordonah, that screen does talk... ACTION_STD_CONTEXT does it
12:00:28 Quit JdGordon (Quit: Leaving.)
12:04:45 Join sasquatch [0] (
12:05:59 Join MethoS- [0] (~clemens@
12:07:11 Quit MethoS- (Remote host closed the connection)
12:07:49 Join MethoS- [0] (~clemens@
12:20:16S_a_i_n_tamiconn: What makes you say that?
12:21:36S_a_i_n_tapps/plugins/pictureflow/pictureflow.c, not apps/plugins/pictureflow.c
12:21:56S_a_i_n_tDoes that make more sense now?
12:22:52S_a_i_n_tcommenting/ifdef'ing out lines 1808 to 1810 fixes the lag issue for me.
12:23:56S_a_i_n_tapps/plugins/pictureflow.c is only ~30 lines, I can understand if that caused confusion :P
12:24:24 Join [Saint] [0] (S_a_i_n_t@
12:24:42 Quit [Saint] (Client Quit)
12:38:05 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:47:52***Saving seen data "./dancer.seen"
12:56:56 Quit tchan (Read error: Connection reset by peer)
12:57:49 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
13:03:55S_a_i_n_tmaraz: Do you mind if I add your real name to thetesting/thanks credits in the iLike theme for Nano1~2G? If so, what is it (PM me if you want)? If you'd rather not...that is cool also.
13:04:08marazS_a_i_n_t: it reads in my whois. ;)
13:04:17marazand no, I don't mind
13:04:42S_a_i_n_tI'm just trying to finish up some details of the theme before I update the theme on the themesite.
13:05:14S_a_i_n_tThanks :D
13:07:04 Quit tchan (Read error: Connection reset by peer)
13:07:54 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
13:17:33amiconnS_a_i_n_t: I do mean apps/plugins/pictureflow/pictureflow.c. The lines you suggest commenting out seem to be completely arbitrary
13:17:58amiconnThe other pictureflow.c is the overlay loader for lowmem
13:18:01S_a_i_n_tamiconn: Try it, that is what the author suggests also.
13:18:15S_a_i_n_tIt made a HUGE difference to the speed.
13:18:38pixelmamaybe that's not the root of the problem then though?
13:19:52S_a_i_n_tWell, as long as it works...and until anyone finds the "real" problem, I'll continue using it.
13:20:15*S_a_i_n_t will find the FS# where the author suggests this is the problem.
13:23:46S_a_i_n_tSorry, long link.
13:24:35 Quit angelwolf71885 (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
13:24:41 Quit liar (Ping timeout: 258 seconds)
13:25:44 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
13:26:03S_a_i_n_tI'd be interested to see/hear of the results on your device with the same code commented out.
13:27:35S_a_i_n_tOn my Nano1~2Gs, with that resync code removed it is still not running 100% compared to what it used to be like...but it is countless times better than the way it runs in current SVN.
13:28:23S_a_i_n_tNo lag, just some slight graphical glitches, such as slight tearing in vertical lines while scrolling fast.
13:39:08 Join kugel [0] (
13:39:14 Quit kugel (Changing host)
13:39:14 Join kugel [0] (~kugel@rockbox/developer/kugel)
14:12:00kugelS_a_i_n_t: the change in pictureflow helped against audio dropouts on the fuze IIUC
14:15:26 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
14:19:38 Quit dfkt (Ping timeout: 265 seconds)
14:22:54 Join wodz [0] (
14:25:33n1skugel: do you feel stronlgy about my proposed fuze keymap change? (as you seem to be the only regular that has voiced specific concerns)
14:26:19kugelno, not strongly. but I'm thinking we should consider the change for e200 too if we want consistency
14:28:12pixelmaand the Clip probably?
14:28:40pixelmaah no, sorry
14:29:16pixelmathat one has no wheel
14:29:50n1sthe e200 keymap seems like a much more infected issue and since i don't have one it feels wrong to argue for changing it
14:31:00JdGordonstop on power for the e200 does make some sense
14:31:14n1sbut i'll go ahead and do it for the fuze then and see if the users will riot
14:31:38n1sJdGordon: yes, i think stop on power is ok
14:35:01kugelpixelma: I think long play is unused on the clip, and stop is on power on it (which is only slightly less akward then on the fuze) so yes on the clip it would be a good idea also
14:36:10JdGordonjust remeber the clip+ has slightly different keys
14:36:15pixelmaI seem to remember I was a bit confused when trying a Clip+ recently
14:36:38JdGordonthe "long play" is up in the lists
14:37:03JdGordonand nothing in the wps
14:37:39pixelmasimilar to the c200... long Play is still stop there. The only thing you lose is being able to stop in lists which you can on some targets
14:38:21CIA-88New commit by mc2739 (r27835): Fix FS #11547 - When selecting radio skin the currently loaded skin is not hilighted when menu is entered
14:40:36CIA-88r27835 build result: All green
14:40:54 Join perfectdrug [0] (
14:41:41 Join fdinel [0] (
14:47:32 Quit Ramsey[LC] (Remote host closed the connection)
14:47:52 Join Ramsey[LC] [0] (~RamseyLC]@
14:47:53***Saving seen data "./dancer.seen"
14:53:55JdGordonanyone want to try a patch which loads the wps and fms on demand instead of on boot?
14:55:22 Join angelwolf71885 [0] (
14:58:33 Quit wodz (Quit: Leaving)
14:59:01JdGordonno noticable wps load difference on my CF-ed mini2g
14:59:27mc2739I'll try it in a bit. link?
15:00:50JdGordonif plugins dont build just do a make bin. It doesnt touch the apis so will work with svn codec/plugins
15:02:01JdGordondont do it with a remote target though... it will crash :)
15:02:12amiconnLoading fms on demand sounds like a very bad idea on hdd targets
15:03:38amiconn(in fact loading anything on demand unless the action requires disk access anyway)
15:03:45 Join hebz0rl [0] (
15:04:47JdGordonI dont have a disk target so unless someone tries it and tells us how bad it is we will never know
15:06:18 Quit angelwolf71885 (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716])
15:06:36amiconnNo need to try and tell. It will cause a ~3 second delay on all 1.8" and 2.5" hdds and the H10 microdrive
15:06:40 Join olejorgenb [0] (
15:07:09olejorgenbthere wouldn't happen to be a sandisk sansa clip+ _4gb_ owner here? :)
15:07:12amiconnIt also means wasting battery if the user enters & exits the radio repeatedly
15:07:24JdGordonno it wouldnt
15:07:27olejorgenbI think I want the 4gb version, but the site only list the 2gb
15:11:57u42polejorgenb: i think i am, let me check
15:13:10u42prockbox runs very well on it :)
15:13:14olejorgenbsweet, then I order that
15:13:25u42pi dont think there is any difference except for the memory being bigger
15:13:57olejorgenbyup, it seems logical
15:14:22 Quit krazykit (Quit: bbl)
15:14:53olejorgenbdo you get about the same battery life with rockbox?
15:15:58u42pno idea, it holds for ages
15:16:29u42pi plug it into my pc every 2-3 days, never had it even nearly 50% i think
15:17:46olejorgenboh, the specs says 15h which is ok, but not ideal for eg. longer bike rides without recharge possibilitis
15:18:02olejorgenbI will do an experiment checking what firmware is best and report back :)
15:18:27u42palso check what codec needs the least cpu
15:19:09 Quit antil33t (Read error: Connection reset by peer)
15:19:16 Join antil33t [0] (
15:20:19u42pi did run that but the output was borked
15:21:22olejorgenbwill do that too
15:23:29 Quit balintx (Remote host closed the connection)
15:25:12 Join [sko] [0] (~sko]
15:25:45 Join balintx [0] (
15:26:25JdGordonis it likely gcc will optimize-out a function call which isnt having its return value used?
15:27:09 Join liar [0] (
15:28:27 Quit perfectdrug (Quit: Leaving.)
15:30:03u42pi have this terrible fear of damaging my hearing or sending bad vibes to my brain when listening to low bitrate music via headphones. it is so crazy
15:30:40n1sJdGordon: if the called function is available when compiling the callee it should
15:31:03n1s(function defined in the same TU (file)
15:31:12JdGordon#gcc is telling me it shouldnt ever happen
15:31:36JdGordonI'm relying on the side effects so the call needs to happen :)
15:32:12n1syeah the compiler will only remove a call if it can prove the function has no side effrects, forgot to mention that
15:32:30n1swhich it can't prove if it doesn't have the fnction in the same TU
15:32:34u42pis there a abx plugin?
15:32:51n1sunless using LTO or if the function is declared with some attrivute IIRC
15:32:58kugelJdGordon: it will do that for "pure" functions
15:33:12n1syeah, pure and const
15:33:20kugelgcc has some heuristics for detecting pure function, maybe it's failing for you
15:33:27JdGordonacronym fail :) I'll just accept a "no"
15:34:12n1sJdGordon: if the function you are calling has side effects (or can't be proven not to) the compiler cannot optimize out the call
15:34:13JdGordon... the code is working fine in the sim so it is being comilied in, just wanted to make sure this is safe for future use
15:38:03 Join jgarvey [0] (
15:38:27kugelmaybe you can just remove the return type, then the function must have side effects
15:38:43 Quit MethoS- (Read error: Connection reset by peer)
15:38:47 Part LinusN
15:38:53JdGordonthat would absolutly break everything else :)
15:39:07n1sit's really nothing to worry about, it's a gcc bug if it deletes calls that has side effects
15:39:57JdGordonI now have a skin manager so instead of each skinned screen managin their own gui_wps structs (And associated crap) it just calls skin_get_gwps(skin, screen) which will load the skin if it isnt loaded already
15:40:18n1swhen are the action preconditions used, they seem arbitrary in the fuze keymap
15:40:29JdGordonso what I'm doing is calling that for each skin + screen on init to make sure they all get loaded :)
15:40:43JdGordonn1s: for long presses and combos
15:41:09JdGordonalso needed on short presses if the button is needed in a combo or long press
15:41:52kugeln1s: I explained them to S_a_i_n_t tonight, should be in the logs
15:42:21n1skugel: what time?
15:42:50n1sfound it
15:44:27JdGordonbtw, ondemand loading should mean we can either shrink the skin buffer, or allow individual skins to be MUCH more image heavy :)
15:44:39 Join MethoS- [0] (~clemens@
15:45:01JdGordontotally wasted on non fm targets right now though :p
15:45:07kugelit means we cannot buffer_alloc it though
15:45:09n1swhy is the precondition even needed for long presses?
15:45:37JdGordonkugel: not necessarily
15:46:23kugelnot? i thought buffer_alloc works only on boot
15:46:46n1salso if it is needed for long presses we have a lot of incorrect entries
15:47:14kugelit's not really needed
15:47:42JdGordonkugel: if playback is stopped then we can use it, and this doesnt change anything there, it still loads the statically sized buffer on boot
15:47:58JdGordonwe have more flexibility with this because we can intellegently manage the skin buffer
15:48:43kugelyou can resize the audio buffer after buffering has been initialized?
15:49:10JdGordonno, but you can reinit audio, cant you?
15:49:32kugeluhm not sure
15:49:36n1swon't that INIT_ATTR stuff bite you then?
15:49:41Tornethis reminds me, buffer_alloc doesn't actually check that it's safe to buffer_alloc
15:49:44Torneand i feel it should
15:49:50Tornebut i'm not sure when it is safe
15:49:58Torneit's safe before audio_init() but is it safe at any time later? :)
15:50:25JdGordonTorne: IIRC pondlife did a patch which would panic if you tried calling it after audio_init() (or something like that)
15:50:38kugeln1s: INIT_ATTR is the smalled problem, someone just needs to remember to remove it if the function is needed later
15:50:39JdGordonbassically yes it is dangerous, but we dont really need to add the protection
15:50:41Tornewell if that's the right logic then i think it should go in :)
15:50:42 Quit watto (Ping timeout: 240 seconds)
15:50:51Tornei mean sure it's not vital
15:51:04Tornebut it's not like it's a lot of code to do one flag check
15:52:12 Quit user890104 (Ping timeout: 248 seconds)
16:03:13JdGordonok, we really should move the apps layer to c++!
16:03:37JdGordonI think this is the 4th fake-OO I've added :)
16:06:14n1shmm stopping playback causes a "pop" in the headphones on the fuze
16:07:03 Join user890104 [0] (~Venci@
16:09:19n1shey cool, fm works while in usb mode :)
16:11:49 Join mamarley [0] (
16:12:12n1swhich remids me of that fm requency drift on the c200...
16:15:13 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
16:20:06 Quit MethoS- (Remote host closed the connection)
16:23:59 Join MethoS- [0] (~clemens@
16:26:36 Join krabador [0] (
16:28:50 Quit MethoS- (Remote host closed the connection)
16:29:11 Join MethoS- [0] (~clemens@
16:30:12CIA-88New commit by nls (r27836): Small change to sansa fuze keymap. ...
16:31:26 Quit mamarley (Remote host closed the connection)
16:31:34*JdGordon kills off that blasted wps_sync_data struct :)
16:32:00CIA-88r27836 build result: All green
16:32:06 Part u42p ("Leaving")
16:32:42 Quit user890104 (Ping timeout: 240 seconds)
16:34:00kugelJdGordon: a skin manager would be nice, it would save a lot of code duplication which each skin has right now because it cannot be done by the skin engine currently (like the forced update when the lcd wakes up)
16:34:52JdGordonthats what my patch does right now...
16:35:54 Join watto [0] (~watto@
16:35:56 Join toffe82 [0] (~chatzilla@
16:36:00JdGordoncall skin_do_full_update(CUSTOM_STATUSBAR, screen) to check if a full update should happen for that skin/screen, or call skin_request_full_update(CUSTOM_STATUSBAR) to make sure both screens do a full update next time
16:41:53 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
16:43:05 Join captainkewl [0] (2669ecc2@gateway/web/freenode/ip.
16:47:57***Saving seen data "./dancer.seen"
16:54:42 Quit JdGordon (Ping timeout: 276 seconds)
17:01:11 Quit Judas_PhD (Quit: Aaarrrggghhh!!!)
17:02:56 Join Judas_PhD [0] (
17:03:32 Quit Judas_PhD (Client Quit)
17:06:25 Join Judas_PhD [0] (
17:15:20kugelthe playback speed difference between my phone and fuzev2 is quite noticeable
17:15:52S_a_i_n_tkugel: The Fuze seems to be the only target I know of presently that benefits from that addition then...everything else (speculation) seems to suffer from it.
17:16:22S_a_i_n_tperhaps it needs to just be ifdef'd for the Fuze?
17:16:47kugelno, it doesn't benefit really, it also runs jerkily should probably just be removed then.
17:17:21S_a_i_n_tremoving it makes my targets run PF at a decent speed again.
17:17:21kugelwhich brings the audio drop outs back :)
17:17:33S_a_i_n_tNot on my targets.
17:17:58kugelthe priority of the loading thread doesn't make any sense to me
17:18:03S_a_i_n_tI don't want my targets to suffer because one has problems with audio.
17:18:37S_a_i_n_tNano1G isn't exactly terribly powerful, and it seems to manage fine.
17:18:43kugelit has a very high priority (and therefore steals cpu time for audio); but most of the time it a) doesn't run at all or b) is blocked by file i/o
17:19:13kugelthe audio drop outs might depend on the performance of the storage driver
17:19:14n1skugel: that sounds bad indeed
17:19:27 Part Zagor
17:19:54kugelI say the yield should be removed and the priority lowered, even if it delays loading slides
17:20:06S_a_i_n_tan idea needs to be formed of what targets exactly actually benefit from the audio dropout/resync code...and which are suffering because of it.
17:20:19S_a_i_n_tNano1 and 2G are definitely better without it.
17:20:55n1sS_a_i_n_t: using a too high priority and then hacking in a workaround for some targets isn't very nice :)
17:21:01kugelhehe, raising the pitch by 1.1% makes the fuzev2 play at normal speed (keeping up with my phone)
17:21:29 Join mamarley [0] (
17:22:26 Quit mamarley (Remote host closed the connection)
17:22:45kugelthe idea of the loader thread was to allow the main thread to render the slides while new slides are being loaded (in another thread, so loading doesn't stall rendering). but the high priority defeats the purpose somewhat
17:23:34n1simo, nothing should be such high priority that it can cause audio to drop out
17:25:35kugelI find it strange though that I can mange to stall audio completely even with very slow scrolling; the audio system can raise its priority on demand, to a higher priority than the pf loader thread
17:26:27 Join funman [0] (~fun@rockbox/developer/funman)
17:26:28kugelI'm thinking that dynamic priority raising doesn't quite work as intended
17:28:44 Part [[sroracle]] ("Exit")
17:30:05 Join mamarley [0] (
17:42:53 Quit hebz0rl (Quit: Ex-Chat)
17:44:59 Join user890104 [0] (~Venci@
17:45:51 Join Topy [0] (
17:46:34 Join komputes [0] (~komputes@ubuntu/member/komputes)
17:48:58 Quit T44 (Ping timeout: 266 seconds)
17:50:40 Join [sko]1 [0] (~sko]
17:51:48 Quit [sko] (Ping timeout: 276 seconds)
17:53:15 Join timccccc [0] (
18:04:35 Join halmi [0] (
18:05:46 Join changeme [0] (
18:06:31 Quit changeme (Client Quit)
18:10:03 Join h3po [0] (
18:16:49 Join [sko] [0] (~sko]
18:18:46 Quit petur (Quit: Page closed)
18:18:56 Quit [sko]1 (Ping timeout: 265 seconds)
18:36:52S_a_i_n_thehehehe...counting black and white, my lamp.rock now has 50 colours :P
18:36:59S_a_i_n_toverkill FTW!
18:38:00n1sS_a_i_n_t: maybe we could let users select custom colors, we have a color picker widget already
18:38:08S_a_i_n_tI'll put the patch up on flyspray...but I have serious doubts about it getting added to SVN
18:38:35S_a_i_n_tHmmm, maybe not...dunno.
18:38:50 Join angelwolf71885 [0] (
18:39:41S_a_i_n_twhite - 48 differing RGB combinations following ROYGBIV - black
18:40:09 Quit h3po (Read error: Connection reset by peer)
18:41:03S_a_i_n_tn1s: There is already colour picker logic for the .colours file
18:41:40 Join hebz0rl [0] (
18:41:54S_a_i_n_tSo, it *shouldn't* be too hard to add...I was going to kludge this big colour patch in with the option to change the brightness on targets with the keys to be able to do so.
18:42:04n1sS_a_i_n_t: yes, that's what i said...
18:42:39n1sidk if anyone cares about custom colours though
18:43:11S_a_i_n_tAh, it is, sorry. I read "if there was..." as opposed to "we have a"
18:43:50CIA-88New commit by funman (r27837): AMSv2: don't reboot on USB plug ...
18:43:51n1sbtw, what is the use case for colorrs in the lamp plugin? or is it just for fun?
18:44:13S_a_i_n_tpersonal preference, fun...pass.
18:44:31S_a_i_n_tI always wondered why there was red, but no other colour.
18:44:34funmann1s: when operating around people using telescopes, red light is preferred as it causes less disruption
18:44:51S_a_i_n_tSo, I went on a mission to see how easy they were to add...turns out the answer is "very"
18:45:00n1sfunman: and in warships!
18:45:17S_a_i_n_tsnipers use red light also.
18:45:20pixelmafunman: red is already there on colour targets
18:45:39funmanpixelma: not only
18:45:45 Join h3po [0] (
18:45:56n1sok, so we have a use case for red, i suppose blue is good for blinding people with too good nightvision? :)
18:46:00CIA-88r27837 build result: All green
18:46:01funmanwon't people think of the mrobe?? :)
18:46:03S_a_i_n_tred is there on non colour targets too!?! :P
18:46:07pixelmaand the M:Robe100
18:46:54S_a_i_n_thehehehe...this is awesome.
18:47:13*S_a_i_n_t wonders how hard it would be to implement changing the brightness also.
18:47:38n1sto be clear i'm not against adding more colors, was just thinking aloud
18:47:49n1sor thinking in text rather
18:47:58***Saving seen data "./dancer.seen"
18:48:07S_a_i_n_tn1s: Oh,'s more a "just for fun thing" and more for me than anyone else ;)
18:48:23S_a_i_n_tI'll put it on the tracker, if anyone else gets a kick out of it, great.
18:48:56pixelmaI can imagine people wanting a few more colours but can't imagine people want to cycle through all of them, so chosing the other colour and using it should stay separate IMO (maybe using up/down to select a new one and so left/right stays changing between white and the other colour
18:49:15pixelmaor something similar
18:49:47S_a_i_n_tgah...keymaps != S_a_i_n_t's territory ;)
18:50:21n1sor implement a sort of virtual grid so you can select colors with all directions
18:50:39n1salso i always wanted Morse signaling in the flashlight
18:51:12S_a_i_n_tOne change I want to add, that is out of my league (which I'm wondering if some kind sould could help me with) is making lamp.rock ignore the idle timeout.
18:51:16S_a_i_n_tIs this possible?
18:51:58S_a_i_n_tI like my idle timeout set low, but its annoying as hell when using lamp.rock
18:52:16pixelmaI'd like one addition on the buttonlight targets - currently the plugin overrides backlight settings to get full brightness or in the case of greyscale/monochrome enabling the backlight if it is turned off, I wished it did the same with the buttonlight (not a brightness setting but enabling it). The buttonlight on the c200 is very bright and I usually turn it off but it probably makes some difference for lamp
18:53:01S_a_i_n_tpixelma: I have no buttonlight targets, but, I had wondered about that also.
18:53:22*n1s hands S_a_i_n_t grep and tells him to use it in apps/plugins and search for "idle poweroff" (for example)
18:53:23pixelmacurrently if the buttonlight is disabled, the plugin doesn't turn it on
18:54:04S_a_i_n_tand seems like a bug.
18:54:27 Quit h3po (Read error: Connection reset by peer)
18:55:05 Join h3po [0] (
18:55:54pixelmajust forgotten probably, I even think I could figure that out but haven't gotten around to it yet
18:56:16 Quit user890104 ()
18:56:30S_a_i_n_tplugins are reasonable new territory for me, so, I'm starting to play around with small things first.
18:56:44S_a_i_n_tbaby steps, baby steps...
19:08:39*pixelma wonders if it would make sense to simulate the buttonlight in the sims too and how complicated that would be
19:10:27funmani don't think it makes sense, but for debugging it would be enough to add a printf() in _buttonlight_on() and _buttonlight_off()
19:11:13pixelmaprobably enough
19:11:26 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:13:05 Join stoffel [0] (
19:16:05pixelmaalthough people already complained about it stating "Hold on/off" and that was removed... maybe something for developing then only
19:17:01S_a_i_n_tspeaking of the sis, "disk access" is something which is mising that drives me insane.
19:17:26S_a_i_n_tand, *missing...even.
19:18:35S_a_i_n_tThe best I can do for testing in the sims is make the disk access icon/animation conditional to hold (or something similar with an on/off value)
19:20:48 Join Jerom [0] (~jerome@
19:32:23 Join halmi_ [0] (
19:33:55 Quit krabador (Quit: Sto andando via)
19:35:23 Quit halmi (Ping timeout: 265 seconds)
19:35:44 Join halmi [0] (
19:37:15 Quit mamarley (Remote host closed the connection)
19:37:52 Quit halmi_ (Ping timeout: 264 seconds)
19:37:53 Join mamarley [0] (
19:38:18mamarleyWhat does it mean when you guys say "s/something/somethingelse"?
19:38:41S_a_i_n_tmamarley: s/mistake/correction/
19:39:06mamarleyThanks! And sorry, I meant to post that in -community but clicked the wrong button.
19:39:12n1sit's a regular sed command!
19:39:13S_a_i_n_t"I have three eyes"
19:41:42 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:44:57 Quit evilnick (Read error: Connection reset by peer)
19:45:13 Quit MethoS- (Remote host closed the connection)
19:45:28S_a_i_n_tHmmmm, IIUC if I want to tell RB that "no, we're not actually idle" I would use "rb->reset_poweroff_timer ();"?
19:45:35S_a_i_n_tOr, is there a better way to do this?
19:47:58S_a_i_n_tI don't want to add a menu option or anything to disable idle timeout, I just want the idle timeout to be disabled while running a particular plugin.
19:48:39 Quit Strife89 (Read error: Connection reset by peer)
19:48:44 Quit funman (Quit: free(random());)
19:54:46 Quit Horscht (Quit: Verlassend)
19:59:16 Join Horscht [0] (~Horscht@xbmc/user/horscht)
20:03:00 Quit mamarley (Remote host closed the connection)
20:05:53 Quit halmi (Read error: Connection reset by peer)
20:05:57 Join Broam [0] (
20:08:52 Join Strife89 [0] (
20:09:36 Quit Horscht (Quit: Verlassend)
20:11:32 Quit angelwolf71885 (Remote host closed the connection)
20:14:05 Quit antil33t (Read error: Connection reset by peer)
20:14:13 Join antil33t [0] (
20:16:25n1sS_a_i_n_t: yes, call that reset function often enough so that the timeout never happens
20:16:54S_a_i_n_t"often enough"?
20:17:02S_a_i_n_tShouldn't once be sufficient?
20:17:14n1snope, it just resets the timer
20:17:24n1slike a button press
20:17:48n1syou can of course just mess with the settings but don't forget to restore them on exit
20:18:07 Join JohannesSM64 [0] (
20:18:10S_a_i_n_tAh...shit, that's a bit beyond me.
20:18:46 Quit stoffel (Remote host closed the connection)
20:18:51S_a_i_n_thow can I make sure that rb->reset_poweroff_timer(); is called again before the timeout expires?
20:19:03gevaertsTemporarily changing settings from a plugin seems like a bad idea to me
20:19:21BroamGenerally it is, simply because you could crash in the middle
20:20:32n1sS_a_i_n_t: call it in whatever loop is waiting for buttons or something, it should have some kind of timeout
20:22:01n1sah, it doesn't have a timeout at all, does that mean the call blocks until a button is pressed or that it retuns BUTTON_NONE immediately?
20:22:08 Quit jgarvey (Quit: Leaving)
20:22:50 Join mamarley [0] (
20:23:12S_a_i_n_tbasically, I just want lamp.rock to run until I exit the plugin...I've *no* idea how to achieve this.
20:23:43BroamS_a_i_n_t: you have a loop that waits for buttons, though - you can change the color with a button press. (assuming lamp.rock is the same one on my fuzev2 install)
20:24:41Broamor is this an interrupt?
20:25:43n1sS_a_i_n_t: it uses button_get(true) which i think blocks until a button is pressed, you should change this to button_get_w_tmo so it times out and returns BUTTON_NONE and thus the loop continues and you can call the reset timer thingy
20:26:28 Quit JohannesSM64 (Quit: WeeChat 0.3.3-dev)
20:26:28gevaertsAdd a timeout to the button_get(), and handle the no button case (e.g. by doing nothing), and call your function after the switch block
20:27:01Broamthat sounds right to me
20:27:17 Part watto
20:27:37S_a_i_n_tPossibly due to the fact that its 7am...but, I just can't see what I'm supposed to be doing here.
20:28:44gevaertsYou need to add exactly three lines of code :)
20:28:44gevaertsSorry, add two, and change one
20:28:44S_a_i_n_tproblem is, adding arbitrary code won't work ;)
20:29:29gevaertsS_a_i_n_t: first of all, add a "case BUTTON_NONE:" somewhere where you think it looks nice and matches the furniture
20:30:17 Quit mamarley (Remote host closed the connection)
20:30:17gevaertse.g. somewhere in the bunch of buttons that's not supposed to do anything
20:31:45gevaertsThen you add rb->reset_poweroff_timer(); right above the "} while (!quit)" line
20:32:02gevaerts(i.e. *after* the switch block, but inside the loop)
20:32:26gevaertsThen you're at the point that every time you press the nonexistent button it resets the timer
20:32:27Broamokay, one logical question - what's the timeout value?
20:32:54gevaertsIIRC the poweroff timer is in minutes, so 50 seconds or so seems to be a safe and not too small value
20:33:11Broamwouldn't it be safer to calculate it from the timer?
20:33:40 Nick storm` is now known as stormehz (
20:33:49BroamI'm not that good at embedded programming.
20:34:37 Quit [sko] (Read error: Connection reset by peer)
20:34:53S_a_i_n_tI'm *really* lost...if some kind sir or madam could pastebin something that "just works", I'd really appreciate it. I'm absolutely willing to admit that I'm out of my depth here...if its not a theme, I'm lost pretty much.
20:35:04gevaertsI'd say not. In the best case it won't make any measurable difference. Remember, this is the lamp plugin, one extra function call and run through the loop once a minute will make *zero* difference
20:36:16 Quit DerPapst (Quit: Leaving.)
20:36:17Broamgevaerts: Right. The entire point of this plugin is to be a lamp: the screen is going to burn more power than a function call
20:36:26gevaertsslightly more, yes
20:36:55gevaertsSo in this case I'd definitely go for the simple solution
20:38:52 Quit CGL (Ping timeout: 276 seconds)
20:38:53gevaertsIf you're lucky it compiles
20:39:12 Join krazykit [0] (~kkit@
20:39:22*gevaerts can't be bothered to test that
20:40:07S_a_i_n_tgevaerts: thanks, I like to learn from looking at examples...I get totally fudged when I'm told "Add X to Y, then do Z"
20:40:28S_a_i_n_tAnd, if it doesn't compile...I'll be looking at my changes, not your I'd say ;)
20:40:45 Quit bieber (Ping timeout: 252 seconds)
20:43:27 Join bieber [0] (
20:47:09 Join halmi [0] (
20:48:00***Saving seen data "./dancer.seen"
20:58:02 Join petur [0] (~petur@rockbox/developer/petur)
21:04:29kugelit looks like there's not much changes needed for a screen resolution independant build
21:04:37 Quit bieber (Ping timeout: 246 seconds)
21:04:46 Join bieber [0] (
21:05:06kugelfor most uses of LCD_WIDTH/_HEIGHT they can expand to a function call/global variable lookup
21:06:41 Quit petur (Quit: reboot)
21:06:47 Join DerPapst [0] (
21:06:54gevaertsAre there actually many of those left in the core? With the UI viewport things are dynamic anyway
21:07:24 Quit t0rc (Quit: Give someone code, help them with one project. Teach someone to code, help them rule the world.)
21:08:34 Join petur [0] (~petur@rockbox/developer/petur)
21:08:34kugel50 in apps/, 400 in firmware (simply grepping LCD_HEIGHT)
21:11:24gevaertsfirmware is going to be full of them. LCD drivers tend to use hardcoded values for some reason
21:12:10 Join mamarley [0] (
21:13:02gevaertsMost of the ones in apps/ (outside of plugins) seem to be irrelevant
21:13:06kugelover 200 of them are in the target tree
21:13:39gevaertsRecording, the debug menu, and the skin buffer size calculation aren't too interesting
21:14:39 Quit h3po (Quit: Leaving.)
21:14:39 Join victortozo [0] (
21:18:07 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
21:20:45kugelgevaerts: many of the places where it needs to be a #define is for setting buffer sizes, buffers which need to be buffer alloc'ed at runtime anyway with SRI (screen resolution independance is way too long)
21:21:31 Join Horscht [0] (~Horscht@xbmc/user/horscht)
21:22:02gevaertsYou're working in a very different environment. You *do* have a working malloc...
21:22:14gevaerts(I assume(
21:22:29kugelbut I'm too lazy to free()
21:23:44*gevaerts threatens to swap kugel out then
21:24:03 Join casainho [0] (~chatzilla@
21:24:17kugelgevaerts: btw, when do you guys decide whether studends have been successful or not (if that didn't happen already)?
21:24:52gevaertskugel: I'd hoped to finalise that yesterday, but unfortunately I hit
21:28:08gevaertsEverything is filled in, and the problematic first question isn't the "pass/fail" one :)
21:29:13kugelbut I guess all questions are equally important for the formal validity
21:29:56gevaertsWell, maybe, but if they don't have any first answer they're going to have to be pragmatic
21:30:19kugel"What a coincidence, all students failed!"
21:30:48*gevaerts promised not to comment on that particular possible outcome!
21:31:30gevaertsFunnily enough one of the questions in the survey is "Did Melange function well as a tool to help you participate in GSoC this year?"
21:31:49kugelyea, I have that too
21:34:36 Quit mamarley (Remote host closed the connection)
21:35:08gevaertsAnyway, a slight delay due to technical reasons isn't going to change anything.
21:40:00 Join Temik [0] (~Temik@
21:40:58 Quit Temik (Client Quit)
21:44:13 Quit GeekShadow (Read error: Connection reset by peer)
21:44:47 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
21:49:55 Quit casainho (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100423141150])
21:54:07 Quit ender` (Quit: printk("autofs: Out of inode numbers -- what the heck did you do??\n"); -- /usr/src/linux/fs/autofs/root.c)
21:57:15 Join mamarley [0] (
21:57:20 Join wodz [0] (
22:07:59 Quit captainkewl (Quit: Page closed)
22:12:37 Quit wodz (Quit: Leaving)
22:13:04 Quit victortozo (Quit: Lost terminal)
22:13:04 Join ender` [0] (
22:15:05 Join BRi7X_ [0] (
22:15:54CIA-88New commit by peter (r27838): Update Dutch language file
22:17:49CIA-88r27838 build result: All green
22:18:11 Quit BRi7X (Ping timeout: 265 seconds)
22:26:55 Quit mamarley (Remote host closed the connection)
22:37:28 Quit bertrik (Read error: Connection timed out)
22:38:09 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
22:41:43 Join MethoS- [0] (~clemens@
22:47:43 Quit Jaykay (Ping timeout: 240 seconds)
22:48:02***Saving seen data "./dancer.seen"
22:53:16 Join piotrekm [0] (~piotrek@unaffiliated/piotrekm)
22:55:23 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
22:58:45 Join webguest689 [0] (
23:01:48 Quit webguest689 (Client Quit)
23:05:26 Join dovshap [0] (
23:07:01CIA-88New commit by mcuelenaere (r27839): Android port: add build script ...
23:08:45CIA-88r27839 build result: All green
23:09:55 Quit n1s (Quit: Lämnar)
23:11:21 Join notlistening [0] (
23:11:47*kugel whishes mcuelenaere would come to irc
23:12:04notlisteningClip+ question do i need to update to bootloader to get USB support? Or does usb not support file transfers in rockbox yet..?
23:12:55 Quit mikroflops (Ping timeout: 276 seconds)
23:13:19CIA-88New commit by mcuelenaere (r27840): Android port: also include native libraries in APK file
23:15:01CIA-88r27840 build result: All green
23:15:02pixelmakugel: would be nice indeed
23:25:14 Quit bertrik (Read error: Connection timed out)
23:26:05 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
23:29:44bluebrotherS_a_i_n_t: I've updated the Theme Editor binaries and started a QtBuildEnvironment to hopefully make it easier to get building running on Windows.
23:31:50bieberIs that a wiki page?
23:33:16bieberbluebrother: Thank you for mentoring this Summer, btw :)
23:36:59 Join mamarley [0] (
23:46:08bluebrotherbieber: yes, I've started that wiki page during the weekend to eventually get a page describing how to set it up (especially for Windows)
23:46:20bluebrotherthere's quite a bit of work left though.
23:46:21 Quit elcan (Ping timeout: 248 seconds)
23:46:54 Quit bertrik (Read error: Connection timed out)
23:46:59bluebrotheroh, and I've got some changes around that allow building an installer of the Theme Editor with a dynamically linked Qt. Need to clean that up and do some more testing though.
23:47:21 Join elcan [0] (
23:47:26bieberThat's great :) Should make it significantly easier to set up a build environment
23:47:38 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
23:48:29bieberbluebrother: Have you filled out my final evaluation/did I pass?
23:49:04bluebrotherindeed. Though using a statically built Qt wasn't necessary at any time −− it's just more convenient when distributing binaries.
23:49:52 Quit petur (Quit: here today, gone tomorrow)
23:49:59bluebrotherbieber: yes, I've filled out the evaluation earlier today. I don't know of any reason not to pass you, so you're passed :)
23:50:11bieberThat's good to hear, thanks :)
23:50:50bluebrotherand thanks for the great work on the Theme Editor (and the low amount of work I had to do, didn't manage to have as much free time as I've wanted)
23:54:33 Join Kitr88 [0] (
23:57:05 Quit Kitar|st (Ping timeout: 264 seconds)

Previous day | Next day