ID:80210
 
Keywords: design, game
Another week spent in game development hell bliss. It's also my last week before academia becomes my new first priority.

Photobucket

Current screenshot - scaled to be a little uglier than the game actually is. Here, the natives are rapidly consuming a Terran base despite the protests of the defensive turrets.


Mindful that I'd probably be using the names of my backups to keep track of the major achievements of the week, I recorded some fairly descriptive ones of what I was up to.

  • Aug 17th: Terran Construction Drones No Longer Just For Show

    Where they were previously just hovering outside of the center of the Terran cities, they now were actually trying to do things. Granted, there wasn't a whole lot for them to do, yet.
  • Aug 18th: Extensive Pathing Work And GUI Debug Support

    You know that A* Pathing I said I added last week? Well, sure, the code was in... but I didn't really force my little units to use pathing yet, until now. Turns out it was really pretty darn far from what I'd call "working" pathing code, even a single obstruction was causing problems.
  • Aug 19th: A Star Is Perfect Enough For Now

    After a day of messing with it, I determined the trouble as that my A* heuristic was busted. It was multiplying the movement costs of the parent steps instead of adding them, causing my units to not so much find a path to where they needed to go as blunder into the right place. Once that was fixed, A* was suddenly remarkably efficient.
  • Aug 20th: More Nitpicking And Beautification In.

    I wanted to get to work on bringing the Terran resource model up, but I couldn't seem to leave the Natives or the A* Pathing alone. Ended up adding breadcrumbs (saved A* pathing information) and a considerable amount of optimization to the Native code. They can now run quite comfortably at about 1000 units.

    "Beautification" doesn't refer to the actual icons, which are still largely placeholders, but rather a process I have in the game which goes through all the turfs and applies overlays to make them look better. The map turfs now looked considerably less blocky.
  • Aug 21st: Big Strides For Terrans

    During my nitpicking days, I had noticed a lot of time was being wasted just running the program again and watching it run until it got to where I suspected I'd find a bug. The game had become too big for that to be a good use of my time. I now instituted a "minimum things should be changed before I can run it again" policy, and it was producing excellent results.

    At this point, a great deal of orders were being passed between the brains of the Terran cities, their construction drones, and Terran command. I now had some turrets up.
  • Aug 22nd: Well The Turrets Spin And Natives Detect Danger Albeit Badly

    From early on, I had some thinking that defensive turrets would be a lot more interesting to fight if they take some time to spin and fire. Today a large part of what I was doing coming up with a formula to convert arctan into an angle that the turrets' Icon is .Turned to make the turret appear to point directly at what they're shooting. I got the angle right, although sometimes they spin in the wrong direction -- it has to do with the flip from -180 to 180 (or vice versa) when something is to the west of the turret - needs tweaking.

    I intended to teach my Natives how to detect if there was a place where a lot of them were dying and avoid it until they could send in some soldiers to clear it out. The heuristic needs a lot of tweaking. It can get really laggy if there's a lot of dangerous spots to keep track of as thousands of natives poll the lists of danger spots. Also, the Natives have a regrettable tendency to ignore danger spots entirely and just charge in and completely annihilate the base. Maybe they know something I don't?
  • Aug 23rd (Today): Terran Resource Backbone Improves

    I switched the Terran cities over to building on an established floor plan instead of the old method of following a list of build orders. I also added some support for proper resource generation and expenditure: it was never the intention that the Terrans would be throwing up walls for free.

    A lot of my Terran resource model is based off of the supply and demand lessons from M.U.L.E. - a bit of a homage to the late Bunten's work - I think the players might appreciate what supply and demand brings to the online game model.
Overall, it was a week of lots of bug fixing followed by some excellent progress once I decided to institute a certain minimum number of changes between compiles.

Though the code I've added is the same that will be used by the players to interface with the game, the current state is actually written from the inside out. The game plays itself - several determined AIs battling it out. I'm nearly to the point where I return to the player level. If I had another week off, odds are pretty good I'd have a playable game to show for it. As it is, it's yet to be determined just how much my homework will impact this endeavor.
Wow, that's starting to look fun.
SuperAntx wrote:
Wow, that's starting to look fun.

It sure getting there. ;) Most of the backbone is pretty much established, now I get to start adding the real "game" parts.

Though I do have a whole third faction I barely touched, they're the least complicated of the three to implement (or so I suspect). Actually, it's pretty interesting I say that, as the third faction is largely how the game started before I decided to take it in a new direction. It'd be like I'm coming back home much wiser for the journey.