SafeArea Behavior is Asymmetric by Team??

[brice]

Hi, this is a question to experienced modders: xander, trickfred, TGF, and anyone else who has played a lot with SafeAreas.

My own tinkering suggests SAs behave differently for teams 0 & 2 versus team 1. I've read through the problems people have had with SAs, and I haven't seen any mention of this. But maybe some past SA problems are the result of this?

I'll describe what I see on linux using 1.4.0b9. If anyone has seen the same or different on another platform, would you please let me know? I plan to use SA machinery in my mod and I'd hate to have to gut it because of platform inconsistencies. (Assuming it ever becomes non-vaporware!)

I tested global, non-dynamic SafeAreas for each Team 0,1,2. If I enter the WorldMap and immediately exit the game, there is always an Offline entry in game.txt for each SA. But if I enter the test level then exit the game, the Online/Offline status behaves differently for teams 0,2 vs. 1. It seems like the inequalities of the comparisons are reversed for team 1. I ran the tests using instant TriffidEggs because they stay put.

Code: Select all

Online/Offline status in game.txt of SAs
after enter level then exit game:

SafeArea:   0             40
unit pop:   0  40     39  40  41
           ------     ----------
team   0:   1   1      0   1   1
       1:   1   0      1   1   0
       2:   1   1      0   1   1
    
hypothesis: for SA = n,
   teams 0   2 : online if pop >= n
   team    1   : online if pop <= n

Does this concur with what you've seen? Thanks!

-brice

[xander]

I think that is what I have seen. Team 0 or 2 SafeAreas are places where you want to move your units (either team 0 green DGs, or team 2 player controllable units). Team 1 SafeAreas are places where there are enemies that you have to destroy. I believe that the "Kill the Ants" objective in Yard uses a team 1 SafeArea.

xander

[brice]

Good point! check the original game. The Yard actually uses both: a team 0 SA=1 to check for Armour production, and a team 1 SA=5 to check for Ant destruction. Interesting that the team 1 SA counts the hidden Anthill populations.

But if the SA comparisons are inequalities (versus exact counts), then there are two situations that can't be detected: rising red populations and falling green populations. I bumped into this asymmetry because I was trying to catch a rising red population.

Inequalities make more sense than exact counts since they'd forgive counting glitches and still provide the behaviors seen in the original game. The Yard examples don't prove this SA behavior is cross platform, but I guess it probably is since IV put a lot of effort into making Darwinia platform independent. I probably should break out my Windows copy to verify some things before I finalize too much.

-brice