Galactose: Diet Edition - Day 5

Today, I finally reached something that can be approximated as a "playable" prototype, albeit, one very early in development.

Overall, things are moving forward surprisingly quickly, and I think I've now finished taming most of the matrix madness for now.

I also managed to get sfx and music, along with a proper chase camera(okay, a very fixed one, but still behind the player!), shots, collisions(very basic), an ai(it wanders) and death, in order to reach a basic first pass at a combat prototype.

I won't claim its "fun" yet, but at least you can fly around and pew pew, although, they don't pew-pew back, *YET*.

One of the more glaring bugs atm, is that all of my models are backwards, which looks *really* odd when looking at the ai fly their hamburger fighters.

Tomorrow, I'll probably add a skybox, try and improve the AI a little bit, and possibly investigate adding some kind of trail effects, and fix some of the remaining bugs. Oh, and probably investigate how to package resources with the .apk so I can have everything self-contained.

Galactose: Diet Edition - Day 4

 

 

Stuff is starting to look right...

Added a simple toon shader effect to the rendering, to more closely match the actual game.

I also got quaternion and the basic stuff for flight controls implemented, and cleaned up a lot of code that had gotten pretty ugly, and finally put everything under version control, yay!

Also updated some of the misc icon textures, etc.

The next step is really probably to get the player actually flying around in a drone, which shouldn't be that hard now, and then implementing trails & shots & collision detection, then some very basic AI, should bring us reasonably close to some kind of graybox style target for tomorrow.

I finally got around to creating this blog today. I apologize for the older posts, I put them there more for reference for the future, and to put this project's timeline into context, most of them are from my facebook posts about this project.

I'm currently trying to find a good way to capture videos of my game on an Android device. I heard about a nifty library earlier, SeeMeGaming, but it appears you have to ask them to get access to it. Other than that, I tried using the webcam on my laptop to record a video of the game running on my cellphone, but that came out pretty poorly, and was pretty uncomfortable. Also, I won't be able to really play using that method, so I'll have to keep looking... 

 Above: Recording footage with a webcam doesn't

work too well

Galactose: Diet Edition - Day 3

Today started off with a lot of annoying things to fix, which I talked about in my "day 2.5" post, but after killing the most offending issues, I've gotten something that is getting reasonably close to an initial graybox.

However, I started doing some benchmarking and unsurprisingly, there isn't a lot of leeway on my phone. I'm hoping to be able to improve performance with batching and other techniques with some more time, but I'm still learning about android in general, and have not yet gotten a good grasp of if I'm CPU or GPU bound, or even really what GPU or CPU (or CPU architecture!!) my phone is running, so I have a *LONG* way to go there. I'm hoping I'll be able to run Intel's GPA, but I have no idea if my phone supports it yet (I'm pretty sure it only works on Intel's phone architecture).

I got a bit sidetracked with performance stuff and did not get around to finishing the flight controls, so thats a big priority for tomorrow, since at the moment all you can do is pan the view around the XZ plane and look at motionless models.

Beyond that, I need to implement collision detection, projectiles and some kind of very basic AI (probably gonna go with flocking for the first pass).

It'll be interesting to see how the massively scaled down battle sizes (I'm suspecting I'm going to be limited to 10-20 total draws or so) will affect the gameplay, etc.

On another note, I'm thinking of setting up a blog and/or a twitter tomorrow since Facebook doesn't feel like quite the right format for these posts...

Though, it is nice to be making this kind of rapid progress on something again.

Galactose: Diet Edition - Day 2.5

Finally fixed the matrix problem, there were actually a few things going on at once, and I just hadn't tried the right permutation of tweaks to fix everything.

Spent some time fighting with an odd issue when normalizing normals and then OpenGL not returning from a draw call when that was done. I'm not sure what to make of it, so I worked around it, though I might try submitting a bug report when I figure out who to submit it to (google? samsung? someone else?).

Above: My MOGA Controller,

They were giving them out for 

free at PAX East 2013

Its not visible from the picture, but I also added MOGA Controller support to it. I'm impressed with what I've seen of the SDK for MOGA - its even easier to get started than even XInput! (Basically, one-two lines of initialization code, then, just poll the controller, though, it also supports an event model, which I haven't looked into yet.

I think my next goal is to actually implement flying and such (ie, 6DOF fun stuff), which shouldn't be too hard once I throw in a Quaternion class...

After that, the next obvious steps are GPU performance testing or basic combat (which involves collision detection, projectiles, trails, and so forth)

Galactose: Diet Edition - Day 2

So, its still very much a WIP, but today I figured out a lot of small things which allowed me to finally render something that looks somewhat close to a cupcake fighter...

Initially, I had to fix some issues with the ObjMeshRenderer, which was mainly that it was using the wrong base for vertices, and that it was only binding one vertex stream for positions.

Next, I had to refactor my resources and other content managing code so that building render stuff was not a giant mess. This worked out pretty well, although, once i get a bit further along I'm probably going to rewrite some of my render glue classes since they're a bit lacking.

I also implemented a very basic touch detector to move the camera forward and backwards, mostly for testing purposes.




 
 Tomorrow, I hope to squash some odd projection issues and toy around with basic flight controls using a MOGA controller, and hopefully investigate rendering performance to get an idea of how many things can be rendered at once, etc. I strongly suspect its going to be far more brutal than even on an old Intel (Laptop) GPU, but I guess we'll see. CPU Performance will also be interesting to look at, I suspect I'm going to be limited to at most a small handful of ships at any one time, at any rate...

Also, I need to start playing around with some trails, but even the LQ trails from Galactose won't work since they still used a bit of geometry expansion in the geometry shader stage, which, is *NOT* present in OpenGL ES 2.0... And thats also going to require blending...

Galactose: Diet Edition - Day 1

Today I started trying to port the indie PC Game I've been working on to my android phone. I started working from a very basic OpenGL test app I had gotten working a few months in the past, but today was the "first" day of actual development work on this "Diet Edition".

The fruits of the day's work can be seen in this image to the left.

Its not much, but its a start.

I wasn't sure at the time, but the deformities in the cupcake fighter mesh are because I had made a mistake in my obj mesh renderer's parsing code.

It was solid red because I had not yet figured out how to pass in multiple vertex attributes so it just had position.

Galactose Diet Edition - Day 0

I started trying out some OpenGL Development on my new android phone today. I didn't make much progress, but I did get to what would be a good starting point later