I allways wanted to make my own graphical engine (DirectX)

Ugh... Probably to late for you to change it now, but DX is a terrible choice for an Indie. You are stuck to Windows only, which means no HumbleBundle, downvoting by Linux lovers on Desura and so on.

I have the main idea in head. But... It's not very funny.. How to add some "attractive" stuffs ? Like mission to achieve, time score, rising levels... ?
What do you think ?

I think it's because you started with an engine. To my experience starting with an engine is a suicide and results in unfun things (at least in my case :D). So, I would reverse it, first think of a game you want to make and then invent an engine that would fit the game the best.

My advice would be quite radical... Throw away all your existing code, then plan a gameplay and code it in Unity3D (as an added bonus you could put a demo on Kongregate since your game would be playable (most likely) via browser).

Well the key to gameplay is getting your core mechanics, core feedback loops in place and building up (and stripping down) from there. So, for some ideas let's look as Euro Truck Simulator which is proving to be quite a popular these days.

One core mechanic is the simulation of the truck itself, handling the controls which is a challenge in execution. Another is the strategic mechanic of route planning.

The game has a classic core feedback loop: You earn money for doing jobs, you spend it on upgrading your truck, or buying a new one. You earn XP which can be spent on skill points to transport more dangerous (and valuable) cargo. Routes unlock, etc.... Classic stuff.

These types of mechanics and feedback loops are found in the majority of games these days, actually. It's not hard to transpose them to a space setting, in fact that exactly describes Elite, Wing Commander: Privateer, Freelancer, Evochron, X-[Beyond the Frontier, 2, 3...]. Of course you're not looking to compete with those titles, but simplify and put your own spin on it, right?

So how about something like this: The player transports cargo between stations, both in space and on planet surfaces. There is a challenge in execution utilizing factors mentioned above, such as gravity, inertia, orbital velocity, weather etc.... The player is rewarded with money for completing these missions based on difficulty (long distance, bad weather, pirate attacks etc....).

You can take a buy low/sell high approach for cargo like Elite, or something simply let the player select what cargo they will carry, and have a direct correlation between risk/reward that way.

Then there could be a strategy layer, plotting a route through a solar system, using/avoiding gravity wells, moons, comets etc....

There could also be combat missions, "go there, kill X pirates" etc... but in my opinion it might be interesting *not* to go that route, instead concentrate on a core mechanic of flying/delivery and have enemy ships be merely one of many challenges on the route.

And then of course you need the core feedback loop, so you should reward the player somehow, such as with ship upgrades, new ships, more solar systems unlocked etc.

It may all sound a little complicated but it doesn't have to be, it can be a very stripped down experience and still be quite fun. And once you find the things that are the most fun you can build upon them.

Anyway, I hope I have helped you get some ideas

Hello Xenobrain,

Thank you very much for your post. You helping so much...

I'm now in holidays, but I will post some resulting ideas latter.

My final goal is a full 4X game with a complete story line, separated in episodes (where each introduces more content).

All the rest of your post suggests some kind of space-sim, not a 4X game.

(umm, main difference being a "commander" of fleets and planets instead of a pilot of a ship)

Check out Kerbal Space Program. The main objective is to visit other planets and gather scientific data. The fun comes from building craft to reach the various planets, adapting to the challenges presented by the differing conditions on each one, and returning the data to your planet of origin in order to unlock new technologies which allow you to build better, more efficient craft. You wouldn't have to go into as much depth as KSP, of course.

A few planets of different sizes, ranging in atmosphere and gravity from almost none to very heavy, a mother ship with a warp drive (unless you're providing the player with an interplanetary single-stage craft, which tends to be a little hard to swallow), and a few landers to cover different entries, landings, and launches (and, presumably, some 0-G manoeuvres) would be enough for a demo. You don't need a mission structure to showcase the technology. If it's a physical simulation, then successful ascent, descent, and rendezvous are the aim of the game, and a lot of the fun would come from the realistic simulation of the cockpit environment. Given a reasonable level of realism, survival would be a reward in itself.