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 2014-04-05

00:18:19 Quit amayer (Quit: Leaving)
00:19:36 Quit Misanthropos (Ping timeout: 240 seconds)
00:26:38 Quit Guest80912 (Ping timeout: 252 seconds)
00:32:36 Quit bertrik (Ping timeout: 240 seconds)
00:39:58 Quit ender` (Quit: You do not need a parachute to skydive. You only need a parachute to skydive twice.)
00:42:42 Join Guest80912 [0] (
01:43:39***Saving seen data "./dancer.seen"
01:48:50 Quit RiD (Quit: A good plan today is better than a perfect plan tomorrow.)
01:54:52 Quit pamaury (Ping timeout: 240 seconds)
01:56:36 Join tertu [0] (~tertu@
02:05:07 Quit lebellium (Quit: ChatZilla [Firefox 29.0/20140331125246])
02:22:05 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
02:46:19 Join Strife89 [0] (
03:00:06 Quit AlexP (Remote host closed the connection)
03:01:45 Quit Rower (Quit: Hmmm...)
03:02:50 Quit tertu (Disconnected by services)
03:02:50 Join ter2 [0] (~tertu@
03:21:36 Quit ter2 (Ping timeout: 240 seconds)
03:41:06 Quit Scall (Ping timeout: 240 seconds)
03:43:43***Saving seen data "./dancer.seen"
03:49:07 Join Scall [0] (~chat@unaffiliated/scall)
04:42:20 Quit amiconn (Disconnected by services)
04:42:20 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:42:22 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:42:53 Quit pixelma (Ping timeout: 240 seconds)
04:43:26 Join pixelma [0] (pixelma@rockbox/staff/pixelma)
04:46:07[Saint]I never really got too bothered about color differences between GIMP/sim.
04:46:32[Saint]I figured that even if I got it right there it was going to look different on *every single color target* anyway. So, meh.
04:47:28[Saint]Even "the same" target can have different LCDs in many cases. I figured it was mostly a losing battle.
05:05:37 Join ter2 [0] (~tertu@
05:35:28soap[Saint], the main problem he's describing isn't actually gimp/sim differences, it's working color vs export color.
05:35:49soap(or should I say the main problem appears to be that).
05:36:23[Saint]Oh. Sure. But I'm saying I was defeated by the realization that even if I got it looking right there as soon as it hit the target all bets were off.
05:38:43[Saint]I suspect that effect would be worsened in this scenario.
05:39:39[Saint]Bending over backwards making sure the sim and GIMP agree on things only to find out it still looks completely different on target and you can't do squat about it.
05:40:27[Saint]Other than perhaps employing expensive color space profiling equipment.
05:42:42*[Saint] Hmmms.
05:43:40[Saint]I wonder if (due to App targets) if we have enough 24bit native targets to warrant detecting this and not downscaling to 16 bit?
05:43:45***Saving seen data "./dancer.seen"
05:44:08[Saint]Though I imagine that might get weird very quickly for resolutions that have both 16 and 24 bit native targets.
05:46:24 Quit ter2 (Ping timeout: 240 seconds)
05:51:59 Join ter2 [0] (~tertu@
05:58:06 Quit TheSeven (Ping timeout: 264 seconds)
05:58:56 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:07:38 Quit ter2 (Ping timeout: 252 seconds)
06:13:03 Join ter2 [0] (~tertu@
06:52:37 Quit Scall (Ping timeout: 240 seconds)
07:27:09 Join Scall [0] (~chat@unaffiliated/scall)
07:35:47 Quit [Saint] (Read error: Connection reset by peer)
07:37:08 Join [Saint] [0] (~saint@rockbox/staff/saint)
07:43:48***Saving seen data "./dancer.seen"
07:45:55 Join |akaWolf| [0] (~akaWolf@
07:56:27 Join Misanthropos [0] (
08:01:24 Join AlexP [0] (~alex@rockbox/staff/AlexP)
08:04:08copper[Saint]: yeah, it looks completely different on the actual device, unfortunately
08:04:42copperthey sure don't have the best of displays
08:06:38copperbut it's annoying to end up with three different values for the same color: one in the 24 bit gimp image, one in the 16 bit BMP export, one in the sim
08:07:09 Join ygrek [0] (~user@
08:12:15copperugh, the iPod Classic display is very blue
08:25:57 Quit ter2 (Ping timeout: 240 seconds)
08:27:48 Quit Misanthropos (Ping timeout: 255 seconds)
08:29:34 Quit KiwiCAM_ (Remote host closed the connection)
08:37:53 Join ter2 [0] (~tertu@
08:48:30 Quit ter2 (Ping timeout: 255 seconds)
08:58:29 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
08:58:41copperI bet it would be possible to do a chromatic analysis of the display with appropriate gear, and then devise a formula to convert RGB values, to compensate for the display's faults
08:59:44 Join ender` [0] (
09:02:25 Quit jhMikeS (Ping timeout: 268 seconds)
09:11:36 Quit mc2739 (Ping timeout: 240 seconds)
09:12:33 Join Scr0mple [0] (~Simon@
09:12:55coppersoap: I think the reason I've been getting better results with exporting to 1555, is because I'm using shades of gray
09:13:13 Quit Scromple (Read error: Connection reset by peer)
09:13:14copperI guess the gimp converts Red, Green and Blue colors individually
09:13:31copperthe red and blue values of the original get converted to the same 1555 value
09:13:52 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
09:14:12copperthe green value however, which is supposed to be the same as the red and blue, is converted to the closest possible value too, but that value will almost always be different than the converted red and blue values
09:14:39copperand the converted green value might be "too far" from the converted red and blue values
09:14:57copperwhat a human would do is, change the red and blue values to be closer to the green value
09:15:32copperbut the Gimp doesn't do that, and the end result is a color that's actually LESS close to the original
09:16:02copperwhen exporting to 1555, all three colors are converted to the same value, which results in a faithful shade of gray
09:16:18copperwhereas the 565 export results in a tinted shade of gray
09:17:17copperyou can see in the rockbox color picker that there is often quite a gap between what's possible to set for green, and what's possible to set for red and blue
09:18:19copperand getting the green value to be as close as possible to red and blue, doesn't work as well as doing the opposite: getting the red and blue values as close as possible to green
09:43:50***Saving seen data "./dancer.seen"
09:48:44pixelmamaybe it would work to set the colour value(s) you want in the Rockbox colour picker, then make a screendump and then use the latter as kind of palette for your theme? I'm not sure though how the screendumps are created on colour displays though, on monochrome displays they really are 1bit so should have the same colour space
09:55:22 Join director [0] (~director@
09:55:58pixelmathat's only fun if there aren't too many values to figure out though
09:56:50 Quit bertrik (Remote host closed the connection)
10:01:35copperpixelma: I tried that, doesn't work
10:01:48coppersomething gets lost in the conversion in the gimp
10:02:11copperright now, exporting to 16 bit 1555 works best
10:02:33pixelmaah, ok
10:02:35 Quit director (Remote host closed the connection)
10:03:29copperanyway, like [Saint] said, maybe I'm banging my head for nothing
10:03:50copperbecause the actual display has its own flaws
10:04:18copperright now I'm adding 3% red and 4% green as an overall tint, in order to get a decent looking shade of gray on the actual device
10:04:32 Join krnlyng [0] (~liar@
10:09:04coppermeh, I'm still not entirely satisfied with the result
10:11:16 Join mowglibv [0] (~mowglibv@
10:13:15 Join SuperBrainAK [0] (~andy@
10:18:42 Join kiwicam [0] (
10:25:40 Quit mowglibv (Ping timeout: 268 seconds)
10:54:43[Saint]copper: really the only answer is spending big bucks on color profiling gear. Load up a test pattern on device, scan it, load up a color profile to suit your exact monitor...yadda yadda...
10:55:17[Saint]Even then you could only do it for the targets that you have access to.
10:56:40[Saint]You're more determined than I, that's certain.
10:57:20[Saint]I came across this myself a while back and just ended up shrugging my shoulders and passing it off as something I couldn't fix.
10:58:02[Saint]I'll steal some of the points I saw you mention earlier regarding exporting, though.
10:58:58[Saint]But lately the targets I deal with seem to have really vivid displays that match my monitors at home very well.
10:59:17[Saint](all app-based targets)
11:10:38 Join rela [0] (~x@pdpc/supporter/active/rela)
11:14:55 Quit [Saint] (Remote host closed the connection)
11:16:15 Join [Saint] [0] (~saint@rockbox/staff/saint)
11:23:53 Quit ygrek (Ping timeout: 240 seconds)
11:30:51 Quit ender` (Quit: It's bad luck to be superstititious. -- Law of Superstition)
11:43:53***Saving seen data "./dancer.seen"
11:47:38 Quit [Saint] (Remote host closed the connection)
11:49:02 Join [Saint] [0] (~saint@rockbox/staff/saint)
12:03:52 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
12:09:30 Nick SuperBrainAK is now known as DormantBrain (~andy@
12:13:11 Quit [Saint] (Remote host closed the connection)
12:14:29 Join [Saint] [0] (~saint@rockbox/staff/saint)
12:24:14 Join lebellium [0] (
12:38:35 Quit [Saint] (Remote host closed the connection)
12:39:59 Join [Saint] [0] (~saint@rockbox/staff/saint)
12:48:55 Quit lebellium (Quit: ChatZilla [Firefox 29.0/20140331125246])
12:49:21 Join lebellium [0] (
13:18:19copperyeah, I'm not surprised that smartphone displays are better
13:18:37coppermy Galaxy Nexus AMOLED display is awesome
13:18:58copperI'm already hating my new theme
13:19:21coppergray sucks
13:19:36coppergray is for losers without talent :P
13:22:05coppersomehow I expect to come up with something Apple-like, in 1% of the time
13:22:11copperWhich is stupid.
13:29:12 Join ygrek [0] (~user@
13:29:34 Quit rela (Read error: Connection reset by peer)
13:31:16 Join stripwax [0] (~Miranda@rockbox/developer/stripwax)
13:32:50 Quit stripwax (Client Quit)
13:40:01 Join ygrek_ [0] (~user@
13:42:33 Quit ygrek (Ping timeout: 240 seconds)
13:43:57***Saving seen data "./dancer.seen"
14:03:21 Quit ygrek_ (Ping timeout: 268 seconds)
14:06:08soap[Saint], this isn't a profiling task _until_ the problem of predictable 888->565 is solved.
14:06:48 Quit Scall (Ping timeout: 240 seconds)
14:07:03soapcopper, have you tried using a different program to do the 24->16 step?
14:07:09soapimagemagick or other?
14:07:14 Join ygrek_ [0] (~user@
14:13:10 Quit AlexP (Read error: Connection reset by peer)
14:13:46 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
14:14:58 Join AlexP [0] (~alex@rockbox/staff/AlexP)
14:15:08 Join Scall [0] (~chat@unaffiliated/scall)
14:29:56 Quit AlexP (Remote host closed the connection)
14:34:02 Join rela [0] (~x@pdpc/supporter/active/rela)
14:36:04coppersoap: hmmm, no
14:36:47copperthat thing is like SoX
14:36:52coppera ton of parameters
14:38:05copperlooking into it
14:38:06 Join ygrek [0] (~user@
14:38:28 Quit ygrek_ (Remote host closed the connection)
14:38:56 Join kugel [0] (~kugel@rockbox/developer/kugel)
14:42:32 Join AlexP [0] (~alex@rockbox/staff/AlexP)
14:45:24copper"But it also shows the biggest drawback of this technique... you do not get any pure gray colors, other than black and white!"
14:45:33coppersounds like what I've been seing
14:46:00 Quit krnlyng (Remote host closed the connection)
14:49:11soapwhere is that quote from, copper ?
14:49:13 Join krnlyng [0] (~liar@
14:49:47soap#gimp is a graveyard. No program I've played with allows >256 colors in a palette-limited workspace.
14:50:37 Quit kugel (Ping timeout: 240 seconds)
14:51:05soapgoogle hurts since searches for "16 bit editing" hits everything because few people verbally distinguish between total bits and bits per channel.
14:52:50copperI don't think it's at all possible to have pure gray with RGB 565
14:54:00copperwhat I don't understand is, what does Rockbox do with RGB 1555 content?
14:54:09copperwhy does that work better?
14:54:39copperI mean I understand why the format itself works better with gray, since it allows for the exact same values for red green and blue
14:54:56copperbut if Rockbox works natively in RGB 565, what does it do with RGB 1555?
14:56:16soapcorrect me where I'm wrong, but 565 should allow the exact same value for r g and b
14:56:32soapthe extra bit gives green twice as much precision, twice as many possible values.
14:56:32coppercheck out the Rockbox color picker in the sim
14:56:42 Join ter2 [0] (~tertu@
14:56:48soapthat's the wrong test.
14:57:19copper"Why are you using a 5,6,5 color map? Normally such colormaps assign the odd color to the blue channel as it is the color our eyes has the least amount of distinction."
14:57:26soapyea, that guy is nuts.
14:57:31copperis he?
14:57:34soapI have never heard of 556
14:57:48soapwell, except when I want two timers in one package.
15:00:19soapI'm a photographer not a scientist, so there may be a whole world I'm missing, but I have never heard of 8 bit images being 556, and have never heard of "hiding" the extra information in the blue channel because it's less perceptible there. That justification he offers is exactly the opposite reason that 565 is the standard, which is /because/ the human eye is more sensitive to green thus we double the precision of the green channel. I tried to read tha
15:00:19soapt user's profile to see if he had a history of making oddball statements but the forum software requires regustration to do that, a task I can't be assed to do.
15:02:11soapcopper, when working with 555 color the MSB gets ignored.
15:02:12 Quit ter2 (Read error: Connection reset by peer)
15:02:27 Join ter2 [0] (~tertu@
15:02:38copperhmmm okay
15:02:59 Quit rela (Read error: Connection reset by peer)
15:03:33copperpixelma: seems like I was wrong: Rockbox displays my 565 export with the correct colors, and a dump results in exactly the same 565 image
15:04:21soapNVM that thought, because we know the issue is in generating a 565 file you like not in displaying a file once generated.
15:04:36copperbut if it's possible to have the same values for red green and blue with RGB 565, why the fuck does the Gimp fuck up the conversion?
15:06:19copperok apparently it's always the green value that's off
15:06:33coppereither larger or smaller than the red and blue values
15:07:36soapok, so that explains my misunderstanding of both the above
15:08:43 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
15:08:52 Join ygrek_ [0] (~user@
15:10:00 Quit ygrek (Remote host closed the connection)
15:10:25copperso the green value is the same, except shifted to the left
15:13:14 Quit ygrek_ (Ping timeout: 255 seconds)
15:14:36coppersoap: why is using the Rockbox color picker the wrong test? It clearly displays all possible values for all three color components
15:15:20copperand when trying to make a shade of gray, again, the green component is always off
15:26:44soapit's the wrong test IMHO because I don't know how it works and there are more direct ways of determining the RGB value of a file.
15:27:09soapit's adding unknowns to the chain.
15:29:19copperwell I did a test with an image with 5 different shades of gray, initially 24 bit
15:29:58copperwhen I export it to rgb 565, the conversion looks bad, but Rockbox displays it as is, with exactly the same rgb 565 values (duh)
15:30:22copperwhen I export it to rgb 1555, the conversion looks good in the gimp
15:30:37copperRockbox then converts it to rgb 565, but does a better job at it
15:30:54copperit sets the green component to something closer to the red and blue components
15:33:32copperokay, when I export to RGB 1555 in the gimp, then load that export in the gimp and export again to rgb 565, I get the exact same values as Rockbox converting 1555 to 565
15:33:52copperin other words, Rockbox and The GIMP convert RGB 1555 to RGB 565 the same way
15:43:59***Saving seen data "./dancer.seen"
15:44:07copperif anyone wants to try my WIP theme on an iPod:
15:44:29copperI keep thinking I should drop the gray and use some colors somehow
15:46:53 Quit Scall (Ping timeout: 240 seconds)
15:54:40 Quit Beta2K (Ping timeout: 265 seconds)
15:55:54copperI'm gonna lower the green a bit
16:00:34 Join PurlingNayuki [0] (uid20107@gateway/web/
16:03:23lebelliumgevaerts: could you add YP-R1 to the home page in Unstable ports?
16:06:34 Join ygrek [0] (~user@
16:10:09 Join Scall [0] (~chat@unaffiliated/scall)
16:10:43 Join Beta2K [0] (
16:13:10 Quit cmhobbs (Remote host closed the connection)
16:26:23 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
16:26:24 Join cmhobbs_ [0] (
16:30:13 Quit rasher (Ping timeout: 252 seconds)
16:32:25 Quit Guest80912 (Ping timeout: 252 seconds)
16:32:37 Quit ter2 (Ping timeout: 240 seconds)
16:33:38 Join ter2 [0] (~tertu@
16:35:12 Quit cmhobbs_ (Remote host closed the connection)
16:35:12 Quit cmhobbs (Remote host closed the connection)
16:36:26 Nick ter2 is now known as tertu (~tertu@
16:36:38 Nick tertu is now known as ter2 (~tertu@
16:41:30 Join rasher [0] (~rasher@rockbox/developer/rasher)
16:43:39 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
16:49:12 Quit Strife89 (Ping timeout: 240 seconds)
16:50:10copperokay, so the process that I've come up with is this:
16:50:43copper1) work on a 24 bit XCF in The GIMP, use all the colors that I like with 24 bit precision
16:51:59copper2) add a red layer on top of it (RGB FF0000) with 2.0% visibility, and a green layer (RGB 00FF00) with 2.7% visibility, to compensate for the iPod Classic's display
16:52:08copper3) export to RGB 1555
16:52:21copper4) load the 1555 export in the gimp, re-export to RGB 565
16:52:50copper5) load the RG 565 export in the Gimp, and finally use the color picker to determine what colors to use in theme files
16:53:21copperwhen that's done, I end up with a set of RGB 565 bitmaps, and theme files that use colors that Rockbox doesn't need to convert
16:53:27copperwhat I see is what I get
16:57:08soapcould probably automate steps 2-5 with imagemagick
16:59:42copperwith reduced green tint:
16:59:52coppernow it's just right, IMO
17:00:40copperit's probably the best I can do, considering that RGB 565 doesn't leave me much room to play with
17:01:12copperhmm yes, much better
17:09:07 Join Rower [0] (
17:19:10 Quit cmhobbs (Quit: Leaving)
17:19:34 Quit Beta2K (Quit: Changing server)
17:19:49 Join Beta2K [0] (~Beta2K@2607:f1c0:841:4eff::20)
17:19:55 Join cmhobbs [0] (
17:19:57 Quit cmhobbs (Changing host)
17:19:57 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
17:20:10 Quit cmhobbs (Read error: Connection reset by peer)
17:20:24 Join cmhobbs [0] (~cmhobbs@fsf/member/cmhobbs)
17:20:57coppernow I just have to doctor my screenshots for the theme site :-/
17:21:27 Join amayer [0] (~amayer@
17:24:13 Part Beta2K
17:27:58 Join tertu3 [0] (~tertu@
17:28:12 Quit ter2 (Read error: Connection reset by peer)
17:36:18 Join Guest80912 [0] (
17:44:00***Saving seen data "./dancer.seen"
18:16:55 Quit amayer (Quit: Leaving)
18:17:00 Join jhMikeS [0] (
18:17:00 Quit jhMikeS (Changing host)
18:17:00 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
18:37:16 Quit dfkt (Ping timeout: 240 seconds)
18:41:08 Quit jhMikeS (Quit: Confucius say: The short dandelion survive the lawnmower)
18:44:33 Quit ygrek (Ping timeout: 268 seconds)
18:46:22 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
19:31:06 Join Misanthropos [0] (
19:44:01***Saving seen data "./dancer.seen"
19:44:25 Quit bluebrother (Disconnected by services)
19:44:30 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
19:44:56 Nick DormantBrain is now known as SuperBrainAK (~andy@
19:51:44 Join ygrek [0] (~user@
19:59:51 Join dimchuck [0] (~dimchuck@
20:01:57 Quit bertrik (Remote host closed the connection)
20:43:20 Quit dimchuck (Remote host closed the connection)
20:56:36 Join dfkt [0] (OxO29A@unaffiliated/dfkt)
21:08:06gevaertslebellium: isn't that a job for whoever cares about that port?
21:16:03lebelliumgevaerts: probably but IIRC editing the page is not enough. There is something to do with the server after so that the change is visible and only you and the swedens can do that, no?
21:16:12gevaertsI can't
21:16:45lebelliumin this case you can't help more than kugel and lorenzo
21:29:35*[Saint] wonders what's going to happen when copper gets "why do the colors look weird on my <insert one of MANY non-Classic 320x240 devices here>?"
21:30:38[Saint]Or, perhaps, even, "why do the colors look weird on this one LCD variant of the two available?"
21:34:10 Quit tertu3 (Ping timeout: 252 seconds)
21:38:47 Quit TBCOOL (Ping timeout: 240 seconds)
21:39:40 Join TBCOOL [0] (
21:44:04***Saving seen data "./dancer.seen"
21:47:53 Quit Scall (Ping timeout: 255 seconds)
21:51:04copper[Saint]: wait, let me look for my careface… I seem to have misplaced it
21:52:14copperseriously though, like every other theme author (probably), I make themes for myself
21:52:27copperI can't have my own theme look bad on my own DAP and not fix it
21:53:36copperso if it looks bad on another target or even another iPod Classic with some slightly different hardware, there isn't much I can do except maybe produce a version without the red and green filters
21:53:51copperbut I'll wait for someone to complain about it before I start maintaining two versions
21:56:34*[Saint] complains about it
21:57:01 Join kugel [0] (
21:57:01 Quit kugel (Changing host)
21:57:01 Join kugel [0] (~kugel@rockbox/developer/kugel)
21:58:16kugeljhMikeS: strlcpy isnt standard. we have string-extra.h for that
21:58:21copper[Saint]: seriously? Did you load it on your iPod Video?
21:59:03[Saint]No. No. Just winding you up. Sorry, I couldn;t resist.
21:59:55copperactually I think there's a chance that other iPods (and possibly other DAPs) are blueish too
22:00:10copperit's not uncommon for cheap displays
22:01:06 Quit ygrek (Ping timeout: 240 seconds)
22:02:17copperanyway, I've come up with just the right amount of red and green for my iPod Classic
22:03:31copperthere's no way I'm gonna change that without making a second version
22:03:31 Quit Guest80912 (Read error: Connection reset by peer)
22:03:32 Join Guest81512 [0] (
22:04:05copperit's too bad that it looks completely off in the sim, of course
22:14:13jhMikeSkugel: thus my last commit message (but strlcpy appears to get declared in some irregular way depending upon the build client)
22:17:09 Join Scall [0] (~chat@unaffiliated/scall)
22:29:44 Quit kugel (Ping timeout: 255 seconds)
22:31:17 Quit krnlyng (Remote host closed the connection)
22:49:35 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
23:02:18 Quit shamus (Ping timeout: 268 seconds)
23:03:06 Join shamus [0] (
23:11:08 Quit Scall (Ping timeout: 255 seconds)
23:37:38 Join Scall [0] (~chat@unaffiliated/scall)
23:39:29 Quit Misanthropos (Ping timeout: 255 seconds)
23:44:09***Saving seen data "./dancer.seen"
23:52:37 Quit TheSeven (Ping timeout: 255 seconds)
23:54:22 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)

Previous day | Next day