Molly Rocket

apple: but what about f2c?

Won — Thu Sep 09, 2010 2:50 pm

Author: Won

Posted: Thu Sep 09, 2010 2:50 pm

Somewhat mooted with this revision:

http://www.apple.com/pr/library/2010/09/09statement.html

Good news (although I still have a distaste for them).

Duke Nukem Forever

blaineallenbrown — Mon Sep 06, 2010 3:02 pm

Author: blaineallenbrown

Posted: Mon Sep 06, 2010 3:02 pm

It seems that the challenge is back on!

http://www.destructoid.com/yup-gearbox-brings-duke-nukem-forever-to-pax-183058.phtml

and

http://www.gamesradar.com/xbox360/duke-nukem-forever/preview/duke-nukem-forever-pax-2010-hands-on-seriously/a-20100904162748561031/g-20100903112053378011

P ≠ NP solved?

jai — Tue Aug 31, 2010 9:39 pm

Author: jai

Posted: Tue Aug 31, 2010 9:39 pm

martinsm wrote:

This is interesting read: http://scottaaronson.com/blog/ (starting from August 9 and going up).

Awesome P != NP for dummies posts. Definitely a good fit for me (dummy)! lol

physics engine for calculating angular momentum?

jai — Tue Aug 31, 2010 9:28 pm

Author: jai

Posted: Tue Aug 31, 2010 9:28 pm

I have been discovering that. I suppose it's not the engines' faults. As I have been trying to mess around with this, I've had to learn a bit more about physics so that I can understand the engines. Even in the most simplified cases, with friction removed, and just doing 2D with basic geometry, it's complex! Much respect to physics people.

Introducing LibVT

julian — Sun Aug 29, 2010 10:05 pm

Author: julian

Posted: Sun Aug 29, 2010 10:05 pm

source is up:

https://sourceforge.net/projects/libvt/files/

public domain truetype font rendering

sean — Thu Aug 26, 2010 4:23 pm

Author: sean

Posted: Thu Aug 26, 2010 4:23 pm

Thanks!

Issues with running the demo

Bob — Thu Aug 26, 2010 12:35 pm

Author: Bob

Posted: Thu Aug 26, 2010 12:35 pm

If someone is still interested, got it running in VS2010, using a Geforce GTX 280.

Stuff I changed:
Removed one file in the project that triggered the custom build thing, it was in glsl shaders I think.
Removed the version line in virt_text.txt
Modified this line in virt_text.txt:
result.grad = vec4(dFdx(address) * gradient_scale, dFdy(address) * gradient_scale);

stb_truetype C# port

Zer — Wed Aug 25, 2010 5:04 am

Author: Zer

Posted: Wed Aug 25, 2010 5:04 am

Howdy all.

I needed a font renderer and tried the Tao Framework FreeType binding for C# today. Five minutes in I find that two of its three 'open font' API calls have the wrong definition. Thinking, 'this can't get any better further in', I decided instead to port stb_truetype to C#.

Most features are tested and work well - I am actively using baked characters and quads, metrics, etc. Types are marked serializable, so you can use BinaryFormatter, etc. to save your baked character coordinates easily (it takes about 30KB for the full Anonymous Pro font included with the sample, on top of the bitmap data). Of course you could alternatively do everything at runtime.

For your enjoyment,
http://zer7.com/software.php?page=truetypesharp

public domain jpeg/png decompressor

SpartanJ — Wed Aug 18, 2010 4:54 am

Author: SpartanJ

Posted: Wed Aug 18, 2010 4:54 am

Sure! No problem, that's the idea!

sophist.h - public domain portable types

sean — Tue Aug 17, 2010 9:45 am

Author: sean

Posted: Tue Aug 17, 2010 9:45 am

Thanks for testing! List updated.

GJK algorithm implementation video

golgobot — Sun Aug 15, 2010 8:14 am

Author: golgobot

Posted: Sun Aug 15, 2010 8:14 am

It goes into some more depth. http://physics2d.com/content/gjk-algorithm

Eyecandy of the day

jai — Tue Aug 10, 2010 12:41 am

Author: jai

Posted: Tue Aug 10, 2010 12:41 am

awesome.

The Monkey Hustle

mariocaprino — Thu Jul 22, 2010 11:15 am

Author: mariocaprino

Posted: Thu Jul 22, 2010 11:15 am

We have just released our first game for the iPhone - and it's now available on the App Store!
http://itunes.apple.com/us/app/the-monkey-hustle/id377368870?mt=8

The Monkey Hustle
It's the classic shell game - where you bet bananas against Emanuel Desperados, the gorilla from the movie Pinchcliffe Grand Prix.
Our goal with the game is:

Present visually impressive animation and graphics
Something to easily pick-up and play for a few minutes at a time
Try to offer something different to Apps currently available

As this is a site for developers I thought I'd mention some of the technical details around the game.
I've been working on the game and it's framework for the last year - trying to get an understanding for iPhone development.

Fast texture uploads are tricky
The game relies on efficiently playing back video in up to 30 FPS.
It turns out the bottleneck when doing fullscreen video in OpenGL ES on iPhone is texture upload speed. Unless you upload textures in the compressed PVRTC, the OpenGL-driver will perform an expensive "twiddle"-operation on your texture.
Throw away that efficient video decoder you coded for another platform; to optimize playback speed your best file format for iPhone is zipped PVRTC textures.

Immediate mode is cool
There are quite a few things that are just easier to code in immediate mode. As an expample I believe most of us are coding our GUI in immediate-style based on Casey's video.
It's therefore sad that OpenGL ES has removed the option to specify vertices in immediate mode. I understand the performance reasons - but as a coder I can be more creative and work more efficiently with an immediate API.
Thus I coded my own immediate mode interface for buffering vertex data. I love being able to rotate sprites by calling something similar to glRotatef() before calling my draw method - and in my case the performance impact is negligible.

Function pointers is the state
I've tried my luck on state machines a couple of times - and later walk away from a big mess of code I hope I do not have to revisit at a later point.
Lately I've been thinking of function pointers as the simplest way to represent a state. For me it's a new way to think about state machines - and fits nicely with immediate style coding. Combine the function pointer with a timestamp of when the state was set - and you have a powerfull way to script timed sequences.

Memory pools work
C is fun and fast - but lacks a standard C library for data structures. Luckily there are several good to choose from - among them Sean Barret's library available through this forum.
I started early toying around with Apache Portable Runtime. The complete library isn't lightweight - but it's well written and almost all source files are independent of each other and may be included seperately in your project.
But beware - it uses memory pools for all resource handling! I was initially put off by this requirement, but after forcing myself to spend time with the library I have come to love them. There is something elegant about structuring you code around pools: "Got a memory warning? No problem, I'll just clear my Cache-pool!"

I hope you enjoy our game and recommend it to your friends.
http://itunes.apple.com/us/app/the-monkey-hustle/id377368870?mt=8

Best regards,
Mario Caprino

Carmack: Sparse Voxel Octree

icastano — Thu Jul 22, 2010 9:45 am

Author: icastano

Posted: Thu Jul 22, 2010 9:45 am

sean wrote:

This seems like it might be huge: http://alice.loria.fr/~nivoliev/Seamless/

I've always felt like something like this should be possible, although I don't think I ever managed to work out even the simple premise "if you just make shared chart edges have the same shape in texture space (i.e. just translated and 90-degree rotated), you can avoid artifacts". But it feels like that was where I was always trying to get my head to and failing.

That is pretty cool, I'm glad somebody worked out the details. However, building the initial quad-parameterization is still a challenging task. The global parameterization algorithms work well on paper, but it's not clear whether they are robust enough to be used in a production environment.

Simple, fast DXT compressor

icastano — Thu Jul 22, 2010 9:08 am

Author: icastano

Posted: Thu Jul 22, 2010 9:08 am

Code:

int err = abs(stb__Lerp13(maxe, mine) - i);
err += abs(maxe - mine) * 3 / 100;

Hmm... using ints the error adjustment term will almost always be zero:

32*3/100 < 1

if you don't want to use floats, then you should probably scale the errors by 100.

Code:

int err = abs(stb__Lerp13(maxe, mine) - i) * 100;
err += abs(maxe - mine) * 3;