TheNut 179 Jan 07, 2014 at 03:41

The three major issues he posted are indeed a problem, although they are just inconveniences. I’ve learned to accept that for the greater good that browsers at least offer the ability to become gaming platforms and even much more. Plus you get use to the limitations rather quickly and just work with it. ECMAScript 7.0 has plans to evolve the language from out of the stone age. Overloading, concurrency, real classes.

I personally would not recommend emscripten as an alternative. Last I tried to use it, it was like trying to build a rocket ship to land on the moon. Documentation was equivalent to interpreting hieroglyphics, toolchain could only run on Linux, requires CLang, and special platform handling. You can’t just dumbo drop Win32 or OpenSSL stuff in there and expect a working JavaScript app. Even if you spend the time to familiarize yourself with that, you’re going to be building huge JavaScript applications. All that bytecode is going to translate into JavaScript code. At least with a pure JS framework, the browser is responsible for streaming MP3s, loading PNGs, JPEGS, etc. Compact, efficient, and using native browser functions. It’s a nice idea in principle, but I just don’t think it’s the right way. I’d sooner promote Google’s native client, which would make more sense.

fireside 141 Jan 07, 2014 at 03:23

Seems to mostly complain about type safety, which I agree with. The cloning position thing, it seems like wouldn’t be very hard to work around and passing by reference is kind of a necessity.

In response to reply on Non-AABB question
Alienizer 109 Jan 07, 2014 at 03:05

Yes, I want to grow it in whatever orientation it is. I’m not sure what you mean tho. An AABB doesn’t have 8 points to define its bounding box (like a frustum do), an AABB only has a (float3) min and max. So I’m lost here Reed!!!

In response to Non-AABB question
Reedbeta 167 Jan 07, 2014 at 02:35

Do you want to grow the box while maintaining a fixed orientation? If so, you can just transform the plane vertices into the box’s coordinate system. The box is axis-aligned in its own coordinate system so you can treat it like an AABB and just take min/max of everything.

piradyne 102 Jan 07, 2014 at 00:58

We’ve released version 1.0.4.Alpha.

dazin09 104 Jan 06, 2014 at 03:55


dazin09 104 Jan 06, 2014 at 03:52

Unity3d Is Very Powerful, we currently don’t have any documentation on the game, as it is still in alpha and were still working on it. The French revolution idea we will come back to, its just that at the time, i didn’t know the true power of the Unity Engine, As now me and my partner have gotten better with it we’ve discovered more things that we can do to make the experience better. We hope to get the game on Steam, It will probably be out mid this year(2014).

In response to Project Root - Indie game
keywiz 101 Jan 05, 2014 at 23:27

Here is the right download link

demonixis 102 Jan 05, 2014 at 11:37

It’s a great news ! What is the script language used in the editor ?

Vilem_Otte 117 Jan 04, 2014 at 20:48

True, well I still have somewhere an implementation of single-pass depth-fail shadow volumes using geometry shader. I even tried plausible shadows using penumbra wedges and visibility buffer (but meh, for complex models this introduces tons of issues, yet yields correct penumbra shadows).

As for shadow mapping - it isn’t possible to achieve correct penumbra shadows with them. And for most solutions they use jittering (which I personally HATE, noise is just pure evil on screen, and players hate it) - so I am using quite heavy filter that doesn’t result in noisy result.

Ray traced shadows are good (I work with ray tracers a lot), doing realtime ray tracing in whitted style isn’t a problem on good gpu. Path tracing is, it is just noisy. Don’t get me wrong, there are ppl with good results (Jacco Bikker with his Brigade for example - nothing against Brigade, it is awesome piece of technology, it’s not just suited for games yet), but it is still noisy and it will still be noisy within next years. I think thats the main problem with path tracing for now, the noise is too huge compared to real world (there is noise too in your vision, but barely visible), I think we need some improvement of current algorithms. Of course, for static frames or movies, it is awesome.

piradyne 102 Jan 04, 2014 at 18:57

My post on frame-based timing in JetEngine.

reduz 106 Jan 04, 2014 at 03:23

Godot UI is really well integrated to the engine, i don’t think it can be separated, as that integration is what makes it so powerful (the whole Godot editorUI is programmed in Godot UI toolkit using the C++ API).

In response to reply on Math problem
Alienizer 109 Jan 04, 2014 at 02:03

I found a solution some place on the net, the problem is exactly the same as the “Ladders in the alley” problem. Here is the way to solve it. I tried to understand the way it’s done, but I can’t. The author is unknown and left no explanation at all (maybe it was Einstein LOL). Do you think you are capable enough to understand it?

a = longest ladder.
b = shortest ladder.
c = ladders crossing height.

Problem = solve distance between ladders at the base.











dist = sqrt(b^2-m^2)
v71 105 Jan 03, 2014 at 23:42

What did you use as gui ? it looks pretty clean and powerfull, can it be extracted and used as a standalone gui ?

In response to reply on Math problem
Reedbeta 167 Jan 03, 2014 at 22:13

There’s three equations in my post, and the three unknowns are alpha, beta, and C. You know A and B. Of course C is the only one you care about, but my point is there are three equations and three unknowns, so it should be solvable. It is not a linear equation or a quadratic one, so it would have to be solved by substitution. You could use trig identities to eliminate the trig functions, expressing cos in terms of tan or vice versa. I tried putting it into Maxima to solve but it did not come up with a result. That might only be because Maxima isn’t smart enough to do it.

In response to reply on Math problem
Alienizer 109 Jan 03, 2014 at 21:47

That’s 3 separate equations, and Stainless’ equation uses 2 unknown and 1 known, unsolvable. Yours is the same, 1 known and 1 unknown, still unsolvable. Each of the 3 equations can’t really be incorporated together can they?. Would the quadratic equation be useful here?

Nathan2222 101 Jan 03, 2014 at 20:47

Ok. Khan will have to do.

In response to reply on Math problem
Reedbeta 167 Jan 03, 2014 at 20:27

You have also C = A*cos(alpha) and C = B*cos(beta). Together with Stainless’ equation C = 10/tan(beta) + 10/tan(alpha), that’s three equations in three unknowns, so should be solvable, in principle.

In response to reply on Math problem
Alienizer 109 Jan 03, 2014 at 19:17

ok, but now we are back to “how to find beta and alpha”?

TheNut 179 Jan 03, 2014 at 16:00

You’ll have to find that out :) Just taking a quick glance, they have linear algebra on there plus some calculus, geometry, and trig stuff. They also cover parametric equations and Cartesian coordinates, which is good. If this is something you’re doing on your own time, then give it a try. Though I personally wouldn’t recommend any online or free courses as a substitute for the real thing.

TheNut 179 Jan 03, 2014 at 15:49

I haven’t used shadow volumes in a long time (since the widespread of shader support in consumer hardware). Shadow casting is one of those problems that currently doesn’t have an ideal technical implementation yet. Both solutions have their pros and cons, but shadow maps give you much more bang for your buck. Easier to implement, smooth looking penumbras, generally lower performance impact (but this can vary on shadow map complexity too), and independent of scene/geometry complexity.

Personally, I can’t wait until real-time ray tracing becomes more of a reality. Working with lights and shadows in a ray tracer is vastly more intuitive than messing around with hack’ish solutions.

benrawlesmusic 101 Jan 03, 2014 at 15:07

Excited to start working on some great games in 2014! Send me a message :)

reduz 106 Jan 03, 2014 at 12:22

Sorry, I realized this does not publish my email. Please write to: juan {at} okamstudio {dot} com.

legione_ 102 Jan 03, 2014 at 10:39

It sounds great! Especially for linux users like me! How can I get in touch with you? Just use the contact form on your website?

In response to Math problem
Stainless 151 Jan 03, 2014 at 08:04

Ok this is just thinking aloud… but

d = A sin (alpha);    // d is the third side of the triangle formed by A and C, alpha is the angle between A and C
e = B sin (beta);      // e is the third side of the triangle formed by B and C, beta is the angle between B and C
c1 = 10 / tan(alpha); // c1 is x coordinate of the intersection point
C - c1 = 10 / tan(beta);

so ….

C = 10/tan(beta) + 10/tan(alpha);

That should be enough to get you started

Welcome to DevMaster, a community-driven game development website of posts and resources!

Recent Tags

indie × 4
game-development × 4
ios × 3
android × 2
mobile × 1
native × 1
macos × 1
physics-engines × 1
windows-phone × 1
sound × 1
music × 1
networking × 1
testing × 1
multiplayer × 1
game-programming × 1
design-patterns × 1
3d-engine × 1
shaders × 1
gaming × 1
cross-platform × 1
royalty × 1
game-design × 1
game-industry × 1
graphics × 1
mmo × 1