Warzone 2100 Forums

I can't understand the use of the "towers". A tower seem to be the same thing than a hard point with holes and less body points.

Some advices ?

a tower can fire above the concrete wall.

you place it like this :

* = bunkers and emplacements
+ = concrete hardpoints
! = towers
- = normal concrete walls


  This side to the enemy

* * * * * * * * * *
* * * * * * * * * * *
-+-+-+-+-+-+-+-+-+-
! ! ! ! ! ! ! ! ! ! ! ! ! ! !


This way you have 3 levels, a low bunker, a med hardpoint and a tall tower, all firing at the same time.

Edit : Yes, i am a turtle...

yes, but a Hard Point can fire above the concrete wall, too.

but not above another hardpoint

Wuh ? The rules have changed ? Hard points can shoot above other Hard points...

ok, but the original developers cheated their way in this...

The idea was to have 4 levels, each with inimpended line of sight from the other

bunker - hardpoint - tower - indirects

that makes sense. But unfortunately they did not develop a true line of sight algoritm so the tower height advantage was lost...

But with watermelon changes, at least, the towers have the advantage of firing from a higher position (and when the ai is patched to refuse to fire when in danger of hitting friendlies, and the combat system is enabled to check friendly fire, the towers will be a must to increase firepower at a wall...

Didn't watermelon jsut tell me today in another thread that friendlies will be exempt from the projectile collision checks?

yes, but he said that this is because current ai cant handle this yet...

Ok. Then, in the future, a LOF need to be free of objects for hit the target ?

If i create a new structure, say a Tower two time higher than normal towers, with normal descriptions in the Stats files, the new tower can fire below the others structure more easily and have a greater range ?

yes, from what the current svn code does... looks like droids are even able to hit a command center in an elevated position from a lower position by just hitting the top of the building (in earlier code they apparently aimed for the base of the building, now the appear to aim at any exposed building part)... looks like some sort of collision detection was added to the projectiles...

but this is all subjective impressions, watermelon knows better...

i can give you a windows compiled svn version if you want to see (but there are still a lot of bugs, not very playable)

JorgeAldo wrote: yes, from what the current svn code does... looks like droids are even able to hit a command center in an elevated position from a lower position by just hitting the top of the building (in earlier code they apparently aimed for the base of the building, now the appear to aim at any exposed building part)... looks like some sort of collision detection was added to the projectiles...

but this is all subjective impressions, watermelon knows better...

i can give you a windows compiled svn version if you want to see (but there are still a lot of bugs, not very playable)

WZ uses raycasts(1-360 rays per cast) to test line of sight/visibility,I didnt change that iirc.

btw what are those bugs that make the svn 'not very playable'?

Because it stops back to desktop randonly (or aparently randonly) leaving a huge stderr.txt file

JorgeAldo wrote: Because it stops back to desktop randonly (or aparently randonly) leaving a huge stderr.txt file

hmmm... zip up the file and send it along....

towers are also cheaper than hardpoints, which is another reason to use them as a row behind hardpoints instead of having a second row of hardpoints. on a side note, if you change that diagram to


*********************
*********************
+++++++++++++++++++++
! ! ! ! ! ! ! ! ! ! !


then you end up spending a little bit more, but trucks are able to repair all your rear defenses safely (cheaper than having to rebuild).

Watermelon wrote: WZ uses raycasts(1-360 rays per cast) to test line of sight/visibility,I didnt change that iirc.

btw what are those bugs that make the svn 'not very playable'?

that sounds extremely expensive for what is necessary. if raycasting was the way to go, it'd be much easier to just find the enemy units and structures within maximum range and raycast against those, working inward from enemies at maximum range (so that if it finds 3 enemies along a single ray, it can fairly well invalidate two of those as possible targets -- the ai can be pragmatic and not even attempt to hit occluded enemies).