Administrator
|
On 10/27/2012 06:11 PM, Pixelapp [via jogamp] wrote:
> "I didn't know that ffmpeg could be used to perform screencasts, thanks for > the tip." > > Me neither. > > Would some one be kind enough to point me to a code snippet using this ffmpeg > screen cast? Attached is my shell script to produce a screencast, i.e. capture the desktop (or part of it) and produce an mp4. ~Sven |
Administrator
|
Thanks for the script Sven.
I have just added some NEWT support into JMonkeyEngine 3.0. How are handled touch input in NEWT? Do they simply generate mouse events? Does headless rendering work with NEWT?
Julien Gouesse | Personal blog | Website
|
Administrator
|
On 11/03/2012 01:28 PM, gouessej [via jogamp] wrote:
> Thanks for the script Sven. > > I have just added some NEWT support into JMonkeyEngine 3.0. How are handled > touch input in NEWT? Do they simply generate mouse events? > We don't differentiate touch or pointer. We support multiple concurrent pointer events (multi-touch), which is currently only impl. for Android. The demo TextureSequenceCubeES2 (line 107) shows how it can be used for a 2-finger zoom: <http://jogamp.org/git/?p=jogl.git;a=blob;f=src/test/com/jogamp/opengl/test/junit/jogl/demos/es2/TextureSequenceCubeES2.java;h=9217e2b5318336bdc17d60fd52e83efb2f922f6f;hb=HEAD#l107> > Does headless rendering work with NEWT? You mean w/o AWT, i.e. that -Djava.awt.headless=true ? Ofc. ~Sven signature.asc (907 bytes) Download Attachment |
Administrator
|
Yes I meant with java.awt.headless set to true.
Julien Gouesse | Personal blog | Website
|
Administrator
|
In reply to this post by Sven Gothel
How can I make a distinction between an event coming from a touch screen and an event coming from a real mouse? Does an event coming from the mouse always have a pressure set to zero?
Julien Gouesse | Personal blog | Website
|
Administrator
|
On 11/04/2012 12:45 PM, gouessej [via jogamp] wrote:
> How can I make a distinction between an event coming from a touch screen and > an event coming from a real mouse? Does an event coming from the mouse always > have a pressure set to zero? That is a good question, right now I don't know. On Android we only have either or machines set-up right now and sadly I have non X11 machine avail w/ touch. At least I could add a 2nd mouse/pointer I guess to test multi pointers on X11 .. etc. Maybe it is a good idea to have something like 'enum PointerType getType(intindex)', if the NEWT driver can differentiate it. Don't know how at this point thought. IMHO best would be to have a real world use case w/ machine and see how that goes. ~Sven signature.asc (907 bytes) Download Attachment |
Administrator
|
In reply to this post by gouessej
On 11/04/2012 04:26 PM, Sven Gothel wrote:
> On 11/04/2012 12:45 PM, gouessej [via jogamp] wrote: >> How can I make a distinction between an event coming from a touch screen and >> an event coming from a real mouse? Does an event coming from the mouse always >> have a pressure set to zero? > > That is a good question, right now I don't know. > > On Android we only have either or machines set-up right now > and sadly I have non X11 machine avail w/ touch. > > At least I could add a 2nd mouse/pointer I guess to test multi pointers on X11 .. etc. > > Maybe it is a good idea to have something like 'enum PointerType getType(intindex)', > if the NEWT driver can differentiate it. Don't know how at this point thought. > > IMHO best would be to have a real world use case w/ machine and see how that goes. > /** * @return the pointer id for the data at index. * return -1 if index not available. */ public int getPointerId(int index) { ... If MouseEvent is constructed via the single pointer constructor, i.e. no array of pointer [currently: X11, Windows, ..] then the pointerID of index 0 is -1 - even though index 0 does represent the single pointer count. If MouseEvent is constructed via the multiple pointer constructor, i.e. w/ array of pointers [currently: Android] then the pointerID of each index [0..count-1] will be >= 0. Integer pointerID's semantic is machine/OS dependent. ~Sven signature.asc (907 bytes) Download Attachment |
Administrator
|
How can I check this semantic under Linux but with a touch screen without testing it with a real tablet? Which native API call is done under the hood?
Julien Gouesse | Personal blog | Website
|
Administrator
|
On 11/04/2012 09:48 PM, gouessej [via jogamp] wrote:
> How can I check this semantic under Linux but with a touch screen without > testing it with a real tablet? Which native API call is done under the hood? You would need to change the NEWT X11 implementation to receive more than one mouse event .. or something - similar like on Android. I haven't looked into X11 event handling regarding multiple pointer events, i.e. multi-touch - where 2 events are not sequential but one w/ multiple 'pointers' involved -> 2 finger press/swipe/etc. If we could 'detect' these semantics, then you could create an appropriate MouseEvent (the multi pointer ctor one) .. and dispatch it. ~Sven signature.asc (907 bytes) Download Attachment |
Administrator
|
Ardor3D is going to use my renderer based on JOGL 2.0 very soon, Renanse asked me to make the pull request:
https://github.com/Renanse/Ardor3D/pull/4
Julien Gouesse | Personal blog | Website
|
Administrator
|
On 11/16/2012 06:37 PM, gouessej [via jogamp] wrote:
> Ardor3D is going to use my renderer based on JOGL 2.0 very soon, Renanse asked > me to make the pull request: > https://github.com/Renanse/Ardor3D/pull/4 Hurray! KUDOS! ~Sven > Julien Gouesse > http://tuer.sourceforge.net > http://gouessej.wordpress.com > signature.asc (909 bytes) Download Attachment |
Administrator
|
Merge done by Renanse.
Julien Gouesse | Personal blog | Website
|
I'm curious, are the Ardor guys going the multiple-backend route, or are they eventually going to make
a decision and go with a single backend....have they given any indication what their plans are? Harvey |
Administrator
|
Hi Harvey
Ardor3D allows to use AWT and NEWT, I added some high level canvases based on NewtCanvasAWT and GLWindow. Ardor3D (<=1.0) allows to use shaders even though its architecture is not entirely shader-based unlike JMonkeyEngine 3.0. According to Renanse, Ardor3D 2.0 will only have a shader-based architecture, it won't work at all without shader support but I disagree with this choice. It is probable that we will have several working renderers like in JMonkeyEngine 3.0. This one officially requires shader support but there is still a renderer based on the fixed pipeline.
Julien Gouesse | Personal blog | Website
|
Administrator
|
This post was updated on .
Hi
The user's guide of JogAmp's Ardor3D Continuation is now available here, enjoy I remind you that JogAmp's Ardor3D Continuation is the subset of Ardor3D that I go on developing and maintaining.
Julien Gouesse | Personal blog | Website
|
Great news! I was upset when I found that ardor3d was dead earlier this year.
|
Administrator
|
:) I couldn't give up on this engine. I'm open to suggestions.
Julien Gouesse | Personal blog | Website
|
Personally I don't have the time (nor inclination to do it) but if some JOGL-knowledgable person feels like it the jMonekyEngine team sure could use some help in maintaining the JOGL-renderer in the upcoming 3.1 release
http://hub.jmonkeyengine.org/t/can-we-expect-jogl-as-an-official-render-option-in-jme/33555 |
Administrator
|
Thanks. They want to remove my stuff from JMonkeyEngine 3.1 :(
Julien Gouesse | Personal blog | Website
|
Given that I don't know a lot about JOGL APIs and I haven't actively used jME in several years I will be glad to help with what limited time I have. I do like both the JOGL and jME projects. gouessej - please don't hesitate to ask for help and I'll do what (little) I can.
|
Free forum by Nabble | Edit this page |