This utility is very cool.

Just wondering, is it possible to extend it, or to use a similiar
idea (transforming bytecode / objects stored in native memory)
to implement Structure of Arrays (vs classic OO Array of Structures)
in an elegant way? One of the things that benefit from SoA are
big particle systems where only part of the particle info needs
to be sent to the GPU.

.rex

ps: I'm hiring graphics and tools people:
http://www.linkedin.com/jobs?viewJob=&jobId=1754526
http://www.linkedin.com/jobs?viewJob=&jobId=1754523
http://www.linkedin.com/jobs?viewJob=&jobId=1653942
(I run the Engine team and we use Java + OpenGL)

Hi Michael,

Looking forward to meeting you and Sven at SG2010.
I've been a long time JOGL user and would love share
some experience and understand the future of JOGL from
you guys.

.rex

TV graphics? Could you elaborate on that a little? I wasn't aware that
there're TVs that can run Java, let alone OpenGL. Or did you mean
broadcast graphics content?

.rex

Thanks for the confirmation.

If JavaFX uses JOGL2, then how can JOGL(2) be an abandoned project
by Sun as I've read from some posts around the web? And does JavaFX
maintain its own private branch of JOGL2?

.rex

Hi Bienator, so what does JavaFX use internally? I was under the impression that
it uses JOGL as its rendering layer. I would appreciate any clarification on this matter.

.rex

You might have run into gimbal locks, which is something that happens
when using Euler angles: http://en.wikipedia.org/wiki/Gimbal_lock

E.g. Given an object, you can reach the same orientation using
2 different rotation sequences with Euler angles:

Rotate on Y-axis by 90 degrees, then
rotate on X-axis by -90 degrees.

is the same as:

Rotate on X-axis by -90 degrees, then
rotate on Z-axis by -90 degrees.

Just hold a rectangular object with your hands and do the
above experiment yourself and you'll realise that there're
more than one way to arrive at the same orientation.

.rex

Here's what I do, but it may not be applicable to you because I
manage my own OpenGL rendering thread and do *not* use the
EDT. That is, my OpenGL rendering is separate from the EDT so that
my UI is not blocked by OpenGL calls.

---

I typically issue a Thread.sleep() before it is time to swap buffers,
in order to spend as little time as possible inside the swapBuffer()
call itself.

You will need to measure how much time your frame takes and
calculate the amount to sleep accordingly. For 60fps, each frame
is 16.667ms long. If your frame takes 5ms to render, then you can
sleep for 10ms and wake up just in time for the swap. You need
to give the system at least 1-2ms buffer due to timer inaccuracies,
so the amount to sleep needs to be clamped. On Windows, the
timer can be made go at 1KHz using this piece of code at the
start of your app:



.rex

Alternatively, the documentation for the utility classes should be updated
to state clearly which GL states are affected and that users should be very
careful about restoring their states.

The complexity of the texture object often defeats a newbie, and most people
just end up writing their own wrapper. The utility classes have their use to
get newbies started quickly, but they need to know about the pitfalls and the
best way is put the warnings in the API reference.

.rex

Hey Sven,

(Great to have met you at GDC09!)

In a related problem, how does one create a GL context using JOGL
in a headless system? That is, none of the windowing system comes
into play because in a system like NV's Tesla, there's no video output.

.rex

I met Ken Russell (and some of the JavaFX team) at GDC09. All I can
say is they are busy with JavaFX.

With that said, I do hope that JOGL will continue in the near future.

This is comparing apples with oranges. O3D is a high-level API and doesn't
expose OpenGL directly. JOGL is a low-level library that maps OpenGL 1-to-1.
I've been doing graphics programming for over a decade and I fail to see the
point of this post.

And for people who find it entertaining to continuously throw rocks at
Java/JOGL/LWJGL, maybe you can consider contributing to it since it is all
open-source.

This bug happens on my MBP 2.5GHz, NVidia 8600GM, OSX 10.5.5.

It is easily fixed (temporarily) using Jeff Addison's TextRenderHack.
The code can be found in this forum.

.rex

If you are on a desktop using AGP, then glReadPixels is going to be slow.
It is possible to get decent speed if you are using PCIe. For example, reading
back 640x480 at 60fps is possible on PCIe.

You also need to be careful about the pixel formats. There are only a few
'fast paths' and on Windows, BGRA is the fast path such that the driver
does not have to perform additional swizzling of the RGB components.

If your card/driver supports FBOs, use that. FBOs are much easier and faster
than PBuffers. Note that the ARB has already approved this extension even
though it still has a EXT suffix to the method names.

.rex

Works on my MBP 2.53GHz, OSX 10.5.5, NVidia 8600GM, Java 1.5.

It's very nice and silky smooth. Good work!

Would be even greater if I could control the camera with my mouse :-)

.rex

Hey Ken,

From your demo that uses 'Java2D's new Single-Threaded Rendering OpenGL pipeline'
to showcase JOGL/Java2D interop, I'm wonder if it is possible to drive the GLJPanel
with a separate thread while the rest of the Swing stuff happens on the EDT?

Will the GLJPanel still show its content correctly (without redrawing) if there're overlapping
light-weight components?

My current use case:
- A GLJPanel that runs at 60fps, driven by a separate 'engine' thread.
- A JPanel on top of it that shows other stuff. e.g. chat window etc.

Is this threading model even possible? Will it cause visual artefacts since
both components are driven by separate threads?

Thanks mate!

.rex

That's a good summary of the problems affecting everyone.
The combination of different OS'es, availability of stable drivers,
choice of pipeline (Windows and Macs having their own set
of problems) and even text rasteriser choices make a
complex problem that has no simple answer.

---

There's one thing I'm particularly concerned with, and I'm sure
many of us here share the same feeling too:

Will JOGL ever become part of the JRE distribution?

I'll be perfectly happy with the 5MB extra footprint for all the things
that this makes possible.



.rex

Thanks Ken!

I've followed up with a response at the bug parade.

I suspect the 6uN b12 build is using the old Java 5
image for the tray icon, because I was expecting
to see the new orange version that's been shipping
with Java 6.

Ken,

I've just installed (offline version) of 6uN b12.

+ Happy to see that it has remained at the 14MB bodyweight.
- The user experience of the installer, while no showstoppers, still
feels way to corporate and not consumer-friendly enough.

I ran 4 applets:
1. http://www.waterlogic.com.sg/smallmedia/gallery/dragonflydemo.html
2. https://jogl-demos.dev.java.net/applettest.html
3. http://www.turbulence.org/spotlight/thinking/chess.html
4. http://www.wefeelfine.org/wefeelfine_pc.html

I did the above in sequence, concurrently, closing and restarting
in random order. Very happy to see that:
+ FF remains alive and never crashed;
+ one JVM instance serving multiple applets in different threadgroups
  work as advertised;
+ applet restart is very fast;
+ all applets work perfectly.

Great work!!

Lastly, please, the tray icon of the JRE needs to be redone
for Windows. It looks too lo-fi and I'm unable to make up the
details or even shape of the Java logo. This is important for the
branding success of Java because this 16x16 block of pixels is
what the user is going to see, remember and mentally connect
it to Java.


.rex

Ken, thanks and super glad that you've been working
on critically important parts of desktop Java.

Rehashing an already debated-to-death question:
What is Sun's strategy for dealing with the entirely empty
space of designer-friendly tools (e.g. Flash studio) so
that non-engineers can create media-savvy RIAs on
top of Java? Let's face it, the current crop of UI tools
are designed by engineers for engineers. This approach
is evidently not going to get desktop Java anywhere
near the RIA space where I think Java can and should
shine in. It is almost funny (and sad at the same time) that
Adobe built its Flex solution on top of Eclipse and released
AIR just a few days ago.

---

Back to the Consumer JRE. I've been following it for
some time and glad that it has reached a milestone.
Seems like Build 13 will be a big one to look forward
to for me.

shwup (if you remember it) will be reborn (the experimental
phase is over) in the next few months incorporating all
the cool stuff you've seen and that's just the tip of the iceberg.
These deployment and version control issues have really
been the achilles heel stopping us from going out in
a big way, so rest assured that we'll be trying out the EA
builds internally.


.rex

Hi folks,

Been a while. How's everyone doing?

Here's a question I could use some collective wisdom on.

1. Is it possible to get JOGL working *within* the browser (FF)
without going thru the applet path?

For example, a FF plug-in that grabs the windows handle of
the browser's rendering space and let JOGL create a GL context
out of it.

2. Is it possible to do this via JWS?

If the answer is Yes to both these questions, then it is possible
to allow the app to fill up (and resizable) the browser space *and*
benefit from JWS's version control + JRE management.


.rex

Hi Ken,

I'm looking for a convenient way to disable all OpenGL
ops without having to bracket all GL methods. I'm wondering
if implementing a custom GL object, like DebugGL/TraceGL,
that simply does nothing and returns, will be considered
abuse and has pitfalls..? Will this work when there's no
active context?

Please advise.. Thanks!

.rex

Hi Ken (and colleague cheerie..),

I'm using a nVidia 6600GT and is able
to reproduce it, just like you described.
Thanks for pinging nVidia on this.

On a different note, Apple OSX's recent
Java update (~15th Feb 2007) now crashes
JOGL+AWT apps on exit. I've sent the crash
report to Apple already.

.rex

Speaking of TGA, which is very OpenGL-friendly
because of its flipped Y-axis.. However, suppose
one has to work with JPGs, what is the preferred
way of flipping the loaded image since JPGs are
not flipped on the Y-axis like TGAs.

I have to work with many large JPGs all the time
so I avoid creating an extra copy of BufferedImage
and flipping it in code by flipping the V texture coord
instead but this must be handled carefully..

Any better ideas?

.rex

Hi! I'm not sure if this is the best place
to post this, but since we haven't gotten
good responses from the usual channels
(We want client engineers but keep getting
CVs from server engineers) I figured what
the heck..

So, if you wanna prove that real-time
multimedia desktop applications are possible
with Java (for PC and Mac), then you really
owe it to yourself to take a look:

http://www.jobsdb.com/SG/EN/Job.asp?R=JDBS052104150&20975

Company:
muvee Technologies (http://www.muvee.com) is
recognized internationally as the pioneer
and leader in automatic video production.
We're bundled with HP PCs, Sony cameras,
ATI video cards etc. We have a mobile
version that is embedded into 9 Nokia phones.
We had a 3-page feature in the Jan 2005 issue
of Forbes magazine. Our Chief Publicist's
interview with Chris Pirillo at the recent
CES 2006 can be heard here:
http://www.thechrispirilloshow.com/help/20060110_tania_chew_on_muvee_autoproducer.phtml

Thanks!

.rex

Hi! I'm not sure if this is the best place
to post this, but since we haven't gotten
good responses from the usual channels
(We want client engineers but keep getting
CVs from server engineers) I figured what
the heck..

So, if you wanna prove that real-time
multimedia desktop applications are possible
with Java (for PC and Mac), then you really
owe it to yourself to take a look:

http://www.jobsdb.com/SG/EN/Job.asp?R=JDBS052104150&20975

Company:
muvee Technologies (http://www.muvee.com) is
recognized internationally as the pioneer
and leader in automatic video production.
We're bundled with HP PCs, Sony cameras,
ATI video cards etc. We have a mobile
version that is embedded into 9 Nokia phones.
We had a 3-page feature in the Jan 2005 issue
of Forbes magazine. Our Chief Publicist's
interview with Chris Pirillo at the recent
CES 2006 can be heard here:
http://www.thechrispirilloshow.com/help/20060110_tania_chew_on_muvee_autoproducer.phtml

Thanks!

.rex

JSR-231 now has a unified way of controlling v-sync:

GL.setSwapInterval(int);

.rex

There's a thread about this awhile ago:

Minimal JOGL: http://www.java-gaming.org/forums/index.php?topic=11653.0

I did some investigating and got the
following results.

1. unzip jogl.jar and rezip it without the manifest directory (otherwise it causes some exceptions with pack200)
2. run "pack200 --repack jogl.jar"
3. run "pack200 --modification-time=latest --deflate-hint="false" --strip-debug jogl.pack.gz jogl.jar"

[size=15pt]jogl.jar: 837KB -> 199KB ![/size]

I think this calls for some serious look
into the possibility of having jogl.jnlp
serve the packed version for Java 1.5+.
Ken? 

.rex

Has anyone tried to find out the smallest
possible size for using JOGL?

jogl.jar is currently 837KB. Can this use Pack200
via Webstart?

jogl-natives-win32.jar is currently 56KB.
If I remove jogl_cg.dll which I don't need
then it is 40KB. I doubt this can go much
smaller but does anyone has any idea?

My aim is to reduce the first-run wait that
the user has to go through. If the user is on
a 56Kbps modem, then that's about a
2-minute wait. A 512Kbps ADSL line will
need about 25-seconds.

Using tricks like parsing the class files
and rip out methods that are not used is
not a good idea because you never know
when you will need them again when your
app is updated, without having to re-download
JOGL.

Any suggestions?

.rex

Hi Ken, would you care to elaborate on what you
mean by the above? AFAIK there're still a lot of CPUs
out there that do not have SSE2 and we'll have to
support those and make sure there's an acceptable
level of performance for them.

So my questions are:
1. How is SSE2 used in JSR-231?
2. How huge are the performance differences and what's the context?
3. 'inefficient code that must be generated' ?

Many thanks in advance!

.rex