#jogamp @ irc.freenode.net - 20141203 05:06:31 (UTC)

20141203 05:06:31 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20141202050631.html
20141203 05:06:31 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20141203050631.html
20141203 07:11:31 * [Mike] (~Mike]@anon) Quit ()
20141203 07:19:55 * zzuegg (~zzuegg@anon) Quit (Read error: Connection reset by peer)
20141203 07:22:18 * zzuegg (~zzuegg@anon) has joined #jogamp
20141203 07:30:21 * eclesia (~husky@anon) has joined #jogamp
20141203 07:31:55 <eclesia> good morning
20141203 08:39:42 * monsieur_max (~maxime@anon) has joined #jogamp
20141203 08:43:27 * gouessej (5ee4b442@anon) has joined #jogamp
20141203 08:43:42 <gouessej> Hi
20141203 08:55:19 <gouessej> rmk0: I assume that I have to use library_animations and library_animation_clips for the key frame animations in Collada
20141203 08:55:47 <gouessej> rmk0: "incomplete animation support ": http://wiki.blender.org/index.php/Dev:2.5/Source/Architecture/COLLADA/Coverage
20141203 09:14:38 <gouessej> rmk0: I can use the free model astroboy (under CC BY-NC-SA) to understand how the animations are stored
20141203 09:46:37 <gouessej> rmk0: https://github.com/gouessej/Ardor3D/blob/master/ardor3d-examples/src/main/resources/com/ardor3d/example/media/models/collada/sony/Seymour_anim2.dae
20141203 09:58:58 <gouessej> rmk0: Blender 2.72b seems to be able to play the animation, I've just tested with the model above
20141203 10:26:59 * gouessej (5ee4b442@anon) Quit (Quit: Page closed)
20141203 11:41:59 <rmk0> gouessej: right!
20141203 11:43:36 * doev (~doev@anon) has joined #jogamp
20141203 12:02:19 * xranby (~xranby@anon) Quit (Remote host closed the connection)
20141203 12:25:07 <eclesia> brrr ... dae
20141203 12:42:44 <rmk0> .o.
20141203 12:42:49 <rmk0> might make the leap to java 8...
20141203 12:43:07 <rmk0> java 7 is EOL in a few months
20141203 12:43:16 <eclesia> already ?
20141203 12:43:24 <rmk0> they're estimating april 2015
20141203 12:43:51 <rmk0> i'm quite interested in the new concurrency stuff
20141203 12:44:19 <rmk0> https://docs.oracle.com/javase/8/docs/api/java/util/concurrent/CompletableFuture.html
20141203 12:44:38 <rmk0> monadic!
20141203 12:44:40 <eclesia> will java follow the chrome frenzy with 20releases each year ?...
20141203 12:44:58 <rmk0> that... seems unlikely
20141203 12:45:20 <eclesia> it was ironic
20141203 12:45:25 <rmk0> that... seems unlikely
20141203 12:45:33 <rmk0> /o/
20141203 12:45:37 <rmk0> \o\
20141203 12:45:42 <rmk0> *ahem*
20141203 12:45:44 <eclesia> _o_
20141203 12:47:18 <eclesia> I think I found my blender bug, a quaternion with 0,0,0,0
20141203 12:47:50 <rmk0> how did that happen?
20141203 12:48:13 <eclesia> and also that blender write quaternion is the order : w,x,y,z and my api use x,y,z,w
20141203 12:48:25 <rmk0> mine uses the latter too
20141203 12:50:09 <rmk0> heh
20141203 12:50:18 <zubzub> doesnt just about everybody uses the latter?
20141203 12:50:26 <rmk0> setting the rotation as a quaternion to (0,0,0,0) in the ui results in an object that can't be rotated
20141203 12:51:17 <zubzub> 13:44 < rmk0> https://docs.oracle.com/javase/8/docs/api/java/util/concurrent/CompletableFuture.html
20141203 12:51:23 <zubzub> ever had a look at guava
20141203 12:51:26 <zubzub> listenablefuture?
20141203 12:51:35 <rmk0> briefly
20141203 12:51:50 <rmk0> not long enough to form any sort of opinion
20141203 12:51:56 <zubzub> completablefuture tries to do...
20141203 12:51:58 <zubzub> everything
20141203 12:52:07 <zubzub> and it's all defined in the same class
20141203 12:52:17 <zubzub> it's 3000 loc iirc
20141203 12:52:35 <zubzub> it's usable
20141203 12:52:37 <zubzub> and horrible
20141203 12:53:10 <zubzub> listenable future tries to do only one thing
20141203 12:53:16 <zubzub> to infomr you when it's done
20141203 12:53:24 <zubzub> and it's relatively small
20141203 12:54:22 <eclesia> hm found something odd : the rotation is called 'rotation axis angle', with 4 values . if it's not a quaternion what is it ?
20141203 12:54:43 <rmk0> mostly interested in being able to bind futures together... monads
20141203 12:54:57 <rmk0> have done a lot of programming in that style coming from haskell
20141203 12:55:26 <zubzub> callback chaining?
20141203 12:56:38 <rmk0> that's... probably one way to refer to it
20141203 12:56:52 <zubzub> https://code.google.com/p/guava-libraries/wiki/ListenableFutureExplained#Application
20141203 12:57:23 <rmk0> oh, right
20141203 12:58:02 <zubzub> I'm sure completable future can do it as well
20141203 12:58:28 <zubzub> if you manage to find the right method in the 1000 methods it exposes
20141203 13:00:03 <eclesia> haaa, found my problem. sometimes it's rotation_quaternion, sometimes it's rotation_axis_angle.
20141203 13:00:22 <eclesia> sneaky blender :x
20141203 13:02:14 <eclesia> end of lunch break back to work code :(
20141203 13:03:03 <rmk0> eclesia: urgh... does this mean that some objects have a rotation as a quaternion and others as an axis angle, depending on what's selected in the ui?
20141203 13:04:21 <eclesia> it's the rotation of a bone in an animation
20141203 13:04:44 <eclesia> seems like it's a quaternion when you don't touch it at the beginning
20141203 13:05:03 <eclesia> but it changes to an axis angle if you rotate it
20141203 13:05:21 <rmk0> blech
20141203 13:05:56 <eclesia> or maybe because I displayed axis angle in the interface ...
20141203 13:06:25 <eclesia> erf, I hope isn't true, otherwise there will be euler angles too
20141203 13:07:14 <rmk0> i don't think the blender people really understand separating interface and implementation
20141203 13:07:20 <rmk0> seems like all of their problems come back to this
20141203 13:16:07 <zubzub> isnt blender written in C?
20141203 13:16:26 <eclesia> python ?
20141203 13:16:50 <zubzub> looks like it's C
20141203 13:16:57 <rmk0> i thought it was c++
20141203 13:17:15 <rmk0> but i'm referring to the fact that they constantly seem to leak implementation details, and that causes problems like the above
20141203 13:18:09 <rmk0> like... rather than use the most general form of rotations internally (quaternions) and then exposing derived values on request via some clean api... they end up using all the different representations internally and giving you back whichever one you selected last in the ui
20141203 13:18:09 <zubzub> sounds like they have a problem of bad code
20141203 13:18:27 <rmk0> that too
20141203 13:18:40 <rmk0> have you seen the "api"?
20141203 13:18:51 <zubzub> no thank you, I value my eyes
20141203 13:18:54 <rmk0> hehe
20141203 13:19:07 <zubzub> I already have to deal with the code crap here at work
20141203 13:19:10 <rmk0> it's not an api by any stretch of the imagination... they just expose their internal data structures via python
20141203 13:19:26 <rmk0> those obviously change every minor release
20141203 13:19:26 <zubzub> but python iz teh awesome!
20141203 13:32:51 * jvanek (jvanek@anon) has joined #jogamp
20141203 13:42:57 * zzuegg (~zzuegg@anon) Quit (Read error: Connection reset by peer)
20141203 13:43:35 * zzuegg (~zzuegg@anon) has joined #jogamp
20141203 13:43:44 <bruce-> do you also drink a lot of coffee?
20141203 13:45:39 <zubzub> depends how bored I am
20141203 13:56:42 * zzuegg (~zzuegg@anon) Quit (Read error: Connection reset by peer)
20141203 13:58:30 * zzuegg (~zzuegg@anon) has joined #jogamp
20141203 15:50:09 * zzuegg (~zzuegg@anon) Quit (Ping timeout: 244 seconds)
20141203 15:50:59 * zzuegg (~zzuegg@anon) has joined #jogamp
20141203 16:04:35 * eclesia (~husky@anon) has left #jogamp
20141203 16:05:39 * jvanek (jvanek@anon) Quit (Ping timeout: 244 seconds)
20141203 16:32:54 * luska72 (8182120b@anon) has joined #jogamp
20141203 16:38:55 * jvanek (jvanek@anon) has joined #jogamp
20141203 16:42:15 <luska72> When I make a 300x200 JFrame (attached to opengl) and set it visible, opengl automatically calls glViewport(<int> 0x0, <int> 0x0, <int> 0x12C, <int> 0xB2) where 12C = 300 and B2 = 178. Which isn't what I set the size to. However when I do the same thing with a GLWindow (of the same size), opengl calls glViewport(<int> 0x0, <int> 0x0, <int> 0x12C, <int> 0xC8) where 12C = 300 and C8 = 200.
20141203 16:42:46 <luska72> Does anyone here know why this might be? I can't find anything about this online, but I don't really know what keywords would be best to include in my search.
20141203 16:43:16 <rmk0> hm, i've seen this too
20141203 16:44:29 <rmk0> i think it may be an issue with borders
20141203 16:44:53 <rmk0> http://forum.jogamp.org/Force-a-GLCanvas-size-td2393774.html
20141203 16:45:21 <rmk0> looks like setting the preferred size on the content pane is the only reliable way discovered
20141203 16:47:03 <luska72> this setUndecorated(true) did the trick, but now the graphics which were showing up before are hidden :/. I'll play around with it for a bit and try to get it working. Thanks again for the help :)
20141203 16:48:11 <rmk0> think NEWT is generally better behaved if you're not tied to swing/awt for some other reason
20141203 16:49:48 <luska72> yeah, I'm actually following a tutorial which used JFrame and GLCanvas in the code. I'm trying to make it work in NEWT instead, which is why I noticed the problem in the first place
20141203 16:49:58 <rmk0> right
20141203 17:06:19 * Eclesia (~eclesia@anon) has joined #jogamp
20141203 17:51:00 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20141203 18:04:10 * [Mike] (~Mike]@anon) has joined #jogamp
20141203 18:04:48 * jvanek (jvanek@anon) Quit (Quit: Leaving)
20141203 18:27:14 * monsieur_max (~maxime@anon) has joined #jogamp
20141203 18:37:36 <Eclesia> rmk0: haha it's even worse. you can store in the animation the rotation in all types (quaternion,euler,rotation,...) with different values
20141203 18:38:24 <Eclesia> it's just a list on bezier 1D modifiers which affect any parameter of the bone
20141203 18:38:44 <Eclesia> translation,scale,rotation X,Y,Z,W ...
20141203 18:38:49 <Eclesia> ^^
20141203 18:39:02 <rmk0> /o\
20141203 18:39:55 <Eclesia> when you click 'I' to create a keyframe. it stores the bone infos in the type they are showned on the screen
20141203 18:40:35 <Eclesia> you can mix quaternion and angles in the same motion ^^
20141203 18:41:07 <rmk0> sounds awful
20141203 18:41:21 <Eclesia> awful but powerful too
20141203 18:51:43 * luska72 (8182120b@anon) Quit (Ping timeout: 246 seconds)
20141203 18:56:31 * Eclesia (~eclesia@anon) has left #jogamp
20141203 20:33:36 * doev (~doev@anon) Quit (Ping timeout: 256 seconds)
20141203 21:54:12 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20141204 05:06:31 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20141204050631.html