#jogamp @ irc.freenode.net - 20130407 05:05:56 (UTC)
20130407 05:05:56 -CatOut- Previous @ http://jogamp.org/log/irc/jogamp_20130406050643.html
20130407 05:05:56 -CatOut- This channel is logged @ http://jogamp.org/log/irc/jogamp_20130407050556.html
20130407 05:37:03 * hharrison (~chatzilla@anon) has joined #jogamp
20130407 07:33:44 * hharrison (~chatzilla@anon) Quit (Ping timeout: 245 seconds)
20130407 09:34:10 * [Mike] (~Mike]@anon) Quit ()
20130407 11:08:56 * DemoscenePassiv (~Lutsche@anon) has joined #jogamp
20130407 15:58:58 * hharrison (~chatzilla@anon) has joined #jogamp
20130407 16:40:49 <hharrison> sgothel: about the offset canvas in the netbeans platform, I'm going to guess this is a jogl side issue..agreed?
20130407 16:43:39 <hharrison> Also, got a report of deadlock on the github tracker, was hoping the reporter would take it ot jogamp, I'll transcribe it myself shortly to bugzilla
20130407 16:44:13 <hharrison> originally here, also the netbeans platform (from my inexpert reading of the traces)
20130407 16:44:14 <hharrison> https://github.com/hharrison/java3d-core/issues/9
20130407 16:44:57 <sgothel> "about the offset canvas in the netbeans platform, I'm going to guess this is a jogl side issue..agreed?" OSX ? I ask for a unit test .. hmm
20130407 16:45:40 <sgothel> hmm ..
20130407 16:45:58 <sgothel> so your issue/9 is on a Windows machine .. interesting
20130407 16:46:40 <sgothel> this might not necessarily be a JOGL issue .. lets see .. AWT deadlocks are often caused by doing modification operations on !AWT-EDT
20130407 16:47:20 <sgothel> "J3D-Renderer-1":
20130407 16:47:20 <sgothel> at java.awt.Component.reshape(Component.java:2144)
20130407 16:47:20 <sgothel> - waiting to lock (a java.awt.Component$AWTTreeLock)
20130407 16:47:20 <sgothel> at javax.swing.JComponent.reshape(JComponent.java:4156)
20130407 16:47:20 <sgothel> at java.awt.Component.setBounds(Component.java:2133)
20130407 16:47:53 <sgothel> the setBounds(..) call is done on !AWT-EDT - bug in .. j3D ?
20130407 16:48:48 <sgothel> still there ?
20130407 16:51:59 <hharrison> sorry, back
20130407 16:52:23 <hharrison> Digging in a bit
20130407 16:53:30 <hharrison> as an aside, do you have a preffered way of citing the jogl project, doing a paper submission shortly
20130407 16:55:10 <hharrison> This looks somewhat like the deadlock that made us move away from Java3d, but in a slightly different form
20130407 16:56:01 <hharrison> Oh, wait, I think I see it: he's using an overridden postRender function
20130407 17:00:45 <hharrison> If he triggers a size change in there, the work is going to happen the renderer thread
20130407 17:01:27 <hharrison> One of the subtle API issues of Java3d is that there wasn't a lot of guidance as to what kind of context those functions are called in
20130407 17:06:53 <sgothel> back ..
20130407 17:08:04 <hharrison> Likely this is a Java3d/application problem
20130407 17:08:12 <sgothel> well .. I dunno the impact of j3d API methods, however, no AWT modification shall happen on non AWT EDT .. you know it
20130407 17:08:30 <sgothel> if that method is an AWT specific one .. sure
20130407 17:08:34 <hharrison> Yeah, I can explain the gory details if you care
20130407 17:09:01 <sgothel> not that interested TBH :)
20130407 17:09:14 <hharrison> Not surprised :-)
20130407 17:09:33 <sgothel> just though the j3d is agnostic .. and hence should call all on AWT-EDT .. but - I don't know the details, and I don't really want to know :)
20130407 17:09:51 <sgothel> how to reference JOGL ? hmm .. http://jogamp.org ?
20130407 17:10:37 <sgothel> whats your paper about ?
20130407 17:10:46 <hharrison> yeah, I was going to ref this:
20130407 17:10:47 <hharrison> http://jogamp.org/jogl/www/
20130407 17:11:00 <hharrison> About our opensource sim software
20130407 17:11:13 <sgothel> http://jogl.jogamp.org .. if you like
20130407 17:11:35 <hharrison> With all the old java3d stuff and vecmath stuff gone finally, jogl2 is our only dependency
20130407 17:12:02 <sgothel> nice nice
20130407 17:12:40 <sgothel> hacking XKB now .. for proper NEWT key code/sym/char values incl. non-US layout :)
20130407 17:18:45 <hharrison> whee!
20130407 18:48:04 * [Mike] (~Mike]@anon) has joined #jogamp
20130407 19:11:40 * hharrison (~chatzilla@anon) Quit (Ping timeout: 245 seconds)
20130407 19:12:19 * hharrison (~chatzilla@anon) has joined #jogamp
20130407 19:43:22 <rmk0> 2.0.2-rc20130404 seems to work well everywhere i've tried it
20130407 19:43:48 <rmk0> need to get hold of some nvidia hardware, i think, as much as i dislike them
20130407 19:49:26 * masterzen (~masterzen@anon) Quit (Quit: Au revoir!)
20130407 19:52:32 * masterzen (~masterzen@anon) has joined #jogamp
20130407 20:31:06 <[Mike]> What type of nvidia hardware?
20130407 20:32:20 <[Mike]> I can provide some testing on gtx465 and gt540m
20130407 21:01:28 <rmk0> [Mike]: don't mind really, i don't currently have any at all
20130407 21:01:53 * rmk0 reads up on the gtx465
20130407 21:07:18 <hharrison> Our group runs NV Quadro 600s on a daily basis
20130407 21:07:57 <hharrison> At least for the narrow subset of functionality our simulation uses
20130407 21:12:11 <hharrison> That's 15 or so people testing it daily, I just need to get better about updating jogl more frequently
20130407 21:17:43 * hharrison (~chatzilla@anon) Quit (Ping timeout: 256 seconds)
20130407 21:24:27 <rmk0> if anyone would like to give things a go:
20130407 21:24:28 <rmk0> http://waste.io7m.com/2013/04/07/instructions.txt
20130407 21:24:49 <rmk0> it'll clone the source to my package, build it with maven, and then produce a reports.zip file containing the output of the test suite
20130407 21:25:15 <rmk0> it'll want access to a graphics context (maybe an X server, or windows desktop), but it won't open any windows as it only does offscreen rendering
20130407 21:25:42 <rmk0> it'll need maven to have access to jogamp.org's mvn repository
20130407 21:26:11 <rmk0> feel free to browse through reports.zip to ensure that i'm not funnelling your credit card information to brazil, or something :)
20130407 21:28:38 <rmk0> http://waste.io7m.com/2013/04/07/io7m-jcanephora-1.0.0-documentation/d1.xhtml
20130407 21:28:44 <rmk0> a snapshot of the documentation
20130407 21:30:51 <[Mike]> How is this different than mvn test ? :)
20130407 21:31:05 <rmk0> no different
20130407 21:31:10 <rmk0> \o/
20130407 21:31:26 <[Mike]> Does your package have its own unit tests?
20130407 21:31:30 <rmk0> it does, yeah
20130407 21:31:41 <[Mike]> Ahh, well makes sense then haha
20130407 21:31:53 <rmk0> hehe, sorry, yeah, i've internalized that "mvn -C clean verify" command
20130407 21:32:04 <rmk0> am in the habit of using it all the time
20130407 21:33:44 <rmk0> i forgot to mention that i use 'fossil' to check out source code in that script, but it's also available as a tarball over http if you don't want to install fossil
20130407 21:33:53 <rmk0> http://fossil.io7m.com/repo.cgi/io7m-jcanephora/timeline
20130407 21:34:09 <rmk0> can log in there as "anonymous" and there'll be links to download a snapshot
20130407 21:34:10 <[Mike]> Yeah, that was going to be another question... Actually never heard of fossil.
20130407 21:34:44 <rmk0> it's not too well known
20130407 21:34:56 <rmk0> http://fossil-scm.org
20130407 21:35:52 <[Mike]> Interesting
20130407 21:43:37 <rmk0> i've just been told by someone else that the 'zip' command may be incorrect
20130407 21:43:52 <rmk0> they're not seeing any surefire-reports directories being created... investigating now
20130407 21:50:43 * DemoscenePassiv (~Lutsche@anon) Quit (Ping timeout: 260 seconds)
20130407 21:51:07 <rmk0> false alarm
20130407 21:53:26 <sgothel> good evening - yes, JOGL unit tests would own your machine for ~30 min :)
20130407 21:53:42 <rmk0> yep!
20130407 21:54:01 <rmk0> thankfully these only take a couple of minutes, on average
20130407 21:54:21 <sgothel> .. and since we don't clear the FB .. your credit card information might actually get transfered if underlying windowing system does not clear it either :)
20130407 21:55:00 <sgothel> IMHO it's the wrong security angle to do such things at this high level ..
20130407 21:55:43 <rmk0> i think the 4.3 spec actually contains the word "security" now
20130407 21:56:12 <sgothel> lol .. yeah, in the spirit of strncpy ..
20130407 21:56:17 <rmk0> hehe
20130407 22:15:20 * hharrison (~chatzilla@anon) has joined #jogamp
20130407 23:24:20 * DemoscenePassiv (~Lutsche@anon) has joined #jogamp
20130407 23:49:43 * DemoscenePassiv (~Lutsche@anon) Quit (Ping timeout: 260 seconds)
20130408 03:51:48 * odin_ (~Odin@anon) Quit (Ping timeout: 264 seconds)
20130408 04:04:04 * odin_ (~Odin@anon) has joined #jogamp
20130408 04:41:37 <hharrison> cd ../../a
20130408 04:41:44 * hharrison (~chatzilla@anon) Quit (Quit: ChatZilla 0.9.90 [Firefox 18.0.2/20130206104040])
20130408 04:42:37 * hharrison (~chatzilla@anon) has joined #jogamp
20130408 04:45:04 * hharrison (~chatzilla@anon) Quit (Client Quit)
20130408 05:06:10 -CatOut- Continue @ http://jogamp.org/log/irc/jogamp_20130408050610.html