Elysian Shadows

I can send out invites, but I need e-mail addresses. I haven't really played around with it much yet.

Android is open to C/++ now... The Android NDK has particular focus on OpenGL games. For the UI and 2D canvas rendering, you still need Java though. Still, I was so happy when I found out they'd be allowing pure C/C++ apps. Yeah, the main issue is that the NDK doesn't have access to the APIs for th...

One of the great things about env3d is that it's dead simple to create an applet with it if you use the BlueJ IDE.

Depends. What do you want to create or work towards? Java is somewhat simpler than C++ to learn, and there has been a lot of really cool work done in more recent years in Java (like Minecraft), but it's not the raw powerhouse that C++ is. If you've never programmed before then I'd recommend this: ht...

I don't think you're understandng my problem. Just as an example, imagine that the AI's x and y both start at 0 and the gem's x and y are 201 and 448.The AI would start increasing it's x by 2 because it is less than the gem's x. Once the AI got to 200, 0, it would increase it's x to 2, but then on ...

Awesome, sounds good. With the way you're doing it, it means your movement always has to be tied to the collision distance of an object on screen. It adds complexity to every AI routine you might develop rather than to your collision system. That's not necessarily a bad thing, but it is a choice. Th...

Ginto8 wrote:

EccentricDuck wrote:stuff

learn2sarcasm

Although JIT-type programs could run faster... if the interpretation gets done in hardware. Honestly, that's the only possibility that would make the speed comparable, much less faster.

Lol, my bad.

It has nothing to do with the AI being stupid. It's a simple AI that does what it's designed to do. The issue is that the collision, as you also pointed out, isn't being calculated because of the precision of detection relative to the step size of the AI. It keeps stepping over the object and collis...

But everyone knows that Java/JIT is faster than C/C++ anyway! Especially in the embedded world. ;) Lol, nah, Java will never be faster. The theoretical concept that JIT can optimize on the fly is good and all, and it may win out on the occasional operation, but I'm not naive enough to think it's ac...

Computer Engineers are schooled in half electrical engineering, half computer science. We specialize in embedded platforms, architecture, operating systems, and that small little niche area where hardware and software meet. Unfortunately the job market is just so software-oriented these days, that ...

My problem (at least in my mind) has nothing to do with collision. I'm simple trying to get the AI's x and y to the gem's x and y. The issue is that, so long as you increment more than one base unit every update call there's a chance of overstepping your target (throw floating point values into the...

Nice one, Im amazed that you havent been hired by a big corporation yet. That's exactly where I DON'T want to work. I interviewed with several. You've always striked me as somewhat entrepreneurial - would that be accurate? I ask partially because finding work environments that support an entreprene...

I just had another thought - what's your collision distance with the gem? If you don't do any fancy checks along the movement vector to detect collisions every frame it should be, at the very minimum, half of the minimum movement distance per frame so that you don't skip over the object. X - AI O - ...

@Van-B
That looks very cool. 2D sounds like a really good fit for what isn't yet out there (since Unity seems to cover 3D pretty well for cross-platform game development).

Alright, I've fixed it so that the AI no longer walks past the gem. However, it now gets stuck when the AI is to the left of the gem. What was the fix to the initial problem? My suggestion would be to print the value of player2Pos.x and gemPos.x side by side so you can see if the values are as they...