Page 2 of 2

Re: Mobile Mex (revisited)

Posted: 30 Sep 2011, 13:33
by PicassoCT
we could add a cpu-heat as UI-element. And now the game is interesting, because if you get heat-death, you loose the means to play.

Re: Mobile Mex (revisited)

Posted: 03 Oct 2011, 14:53
by PicassoCT
hoijui wrote:ahh.. this juzza thread gave me idea:
when unit deploys, it drills it pops out a huge gun that shoots dirt.
when the dirt lands, it terraforms the map there, making it a bit higher. as it has to drill down, it sacks into the ground deeper and deeper. so you can block incomming units on sight, or you can let it auto-patrol shoot in a circle around itsself on max range, to create a terraform fortress, or just to kill enemy CPUs through path-finder updating cycles.
what would be reasonable intervall if somebody (looks to the ceeling) had such massive terraforming in mind?

I planned a spaceship that would rip out mountains and throw them on its enemys.. and noaw you telling me its impossible due to pathfinding..

Re: Mobile Mex (revisited)

Posted: 03 Oct 2011, 14:56
by AF
You could always try it? There's already code in the engine to create craters, the same but inverse?

Re: Mobile Mex (revisited)

Posted: 03 Oct 2011, 15:00
by PicassoCT
nah, the problem is how often to change the landscape... ill certainly will try that...

but if a massive landscape change does lead to nervouse breakdown among pathfinding i rather keep those changes in bursts with time to recover in between..

Re: Mobile Mex (revisited)

Posted: 05 Oct 2011, 00:43
by knorke
Problem was..it turned into mex every smal stop.
Most things stay unchanged except for this:

Code: Select all

function script.StartMoving()	
	Signal (1)
	killMex()	
end

function script.StopMoving()	
	StartThread (waitToMakeMex)
end

function waitToMakeMex ()
	SetSignalMask(1)
	Sleep (2000) --   <- adjust this number
	makeMex()
end
Now the unit has to stand still for 2 seconds to extract metal.

Re: Mobile Mex (revisited)

Posted: 05 Oct 2011, 11:37
by Google_Frog
PicassoCT wrote:nah, the problem is how often to change the landscape... ill certainly will try that...

but if a massive landscape change does lead to nervouse breakdown among pathfinding i rather keep those changes in bursts with time to recover in between..
From my most recent testing I guess that an area of 2048x2048 updated every few seconds should run fine.

My test consisted of modifying an entire 16x16 map at once, there was a freeze of about half a second while doing so.

Re: Mobile Mex (revisited)

Posted: 05 Oct 2011, 18:45
by PicassoCT
good information is good! You tested with units walking around?