Screenshots are not really saying much.
(making a nice scene just shows if you know how to set up a nice scene,
or in other Terms: a detailed AK-47 model does not make a good Shooter already)
I judge when I see a demo or at least a througout gameplay video.

The point still stands: first make a smaller project to
get familiar with the engine and workflow.

... I actually plan on making an RTS again, just not shure
if I should use gamestudio again or Java+OpenGl,
It would be more work, but the advantage is it can run on all systems and even directly in a Browser.

And I could keep the whole game (without the libraries) under 100Kb.

I bought A7 a little over a year ago & played around with it on and off.

I've already figured out how to do most of what I will need to do. I currently have about 87% of know how for building this game.

I have at least a basic understanding of the exact code I will have to wright for everything the game will do.

It will be a multi-player RTS space game. If the development process goes smooth enough I may even make it a FPS / RTS but only if everything comes together nice enough.

I will say that building the actual game is not nearly as hard as sticking with one decision. Deciding on how something will look is my biggest challenge.

I'm always 1,000,000 steps ahead of myself during development.

Best of all, I'm having allot of fun facing this challenge.

Not the secure mp part? The shader part? The gameplay mechanics? Or creating all the needed graphics?

Anytime I know exactly what to do I can normally work at a decent pace. Lite-C has many automated features which I plan to take advantage of.

I have the multi-player part well planned out (created a MP code template).

I'll most likely be using the built in shaders.

The gameplay mechanics is well thought out (I make sure I know how to do something from the beginning by building small practice levels).

As far as graphics go, I'll be using purchased textures and will edit them to my needs. The models however (most if not all) will be development by me.

Everything is going smoothly so far although I do expect to run into some minor problems during development.

If you are going for an RTS with many units, you should look into
syncronized simulation for the multiplayer.

Its revolves around the concept of only communicating action-commands to the server and back to all clients.

The game itself must make shure that all clients have receive the
SAME actioncommands every game-logic tick.

Then they execute them based on the exact same (using syncronized seeded Random values) manner.

This is the concept used in the Age of Empire games for example.
It enables you to control several hundred RTS units by only
sending a few bytes of actioncommands.

Its not appicable to fast-reaction MP Games (Shooters/Racing)
but to slower reacting games with many entities.

Here a doc on that:

SyncronizedSimulation

thank you for the link Damocles


,
gri

This one is even better, with more detail to the topic

http://www.gamasutra.com/view/feature/3094/1500_archers_on_a_288_network_.php

nice articles, thanx!
in connection with age of empires I-II, there are some ai related interesting atricles too. if anyone interested, I can give some links...

Just post them, AI papers are always welcome

I will collect some, and post in user resources section... I saved a lot of articles but have to find the links again...