Page 3 of 3

Re: 2c-Highground map mistakes and corrections

Posted: 05 Feb 2014, 07:28
by NoQ
Already found some typos ...
getStructureBounds
Will have a look (:
the sides[][] array could be initialised independently of direction

Re: 2c-Highground map mistakes and corrections

Posted: 05 Feb 2014, 09:35
by Cyp
Looking at that, probably better to do something as unlike the original as possible.

Hmmm, if trying to do it sensibly instead of just in a loop like I was thinking, maybe a tiles[], and std::stable_sort by distance to the model's apparent exit point? (To be exactly symmetric, the exit point shouldn't be exactly on a tile, so that the tiles would have a proper order.)

The newly-created droids should probably inherit the factory's orientation, too. Or maybe even inherit the orientation corresponding to the direction from the centre of the factory to where the droid was placed.

Re: 2c-Highground map mistakes and corrections

Posted: 06 Feb 2014, 14:04
by NoQ
I think i'm done :geek:
Push to bugfixes?

I also switched placeDroid() to world coordinates because we aren't really returning a tile (and trying to convert it to a tile makes things much more obscure).

Re: 2c-Highground map mistakes and corrections

Posted: 06 Feb 2014, 20:46
by Cyp
Looks ok, although I think the last ifs bias the droids towards the middle of an edge, such that 3 droids coming out at once would appear with ½ tile spacing (and more droids would come out with the usual 1 tile spacing). Haven't tested, though, might have misunderstood something.

One way of finding the nearest point to a rectangle might be something like »wx = clip(wx, worldCoord(bounds.map.x), worldCoord(bounds.map.x + bounds.size.x));«. And maybe to be extra pedantic, use reallyBuildDroid instead of buildDroid (then can also remove the turnOffMultiMsg(...); calls), and pass it psStructure->rot.

Re: 2c-Highground map mistakes and corrections

Posted: 08 Feb 2014, 08:55
by NoQ
Looks ok, although I think the last ifs bias the droids towards the middle of an edge, such that 3 droids coming out at once would appear with ½ tile spacing (and more droids would come out with the usual 1 tile spacing). Haven't tested, though, might have misunderstood something.
Well, droids instantly move to a rally point anyway (default placement of which, by the way, could also use some rotating). Also, there are less structure vertices than adjacent tiles, so some cluttering is inevitable (pigeonhole problems), unless we actually put all exit points uniformly around the structure, but that would mean that normal exit point would be slightly misplaced, compared to the original game.

Also, for master it would probably make sense to add structure exit point parameter to structure stat files (?) so that eg. a modded factory could produce droids to the right by default :hmm:

Re: 2c-Highground map mistakes and corrections

Posted: 17 Mar 2015, 00:48
by montetank
After long time i played this great map again. Nice changes. But why is that still possible? (screenshot2). It should be better to enlarge the gateway by only 1 pix. (Screenshot 1-i changed for myself) . It has no influence on the balance, and this stupid blocking against Nullbot wouldn`t be possible anymore. In the game i have 12 oils and Nullbot was sitting at home with 2 oils. :? What is the reason, that this wasn`t changed?

Re: 2c-Highground map mistakes and corrections

Posted: 17 Mar 2015, 03:12
by vexed
montetank wrote: It should be better to enlarge the gateway by only 1 pix. (Screenshot 1-i changed for myself) .
I assume you mean 1 tile, not 1 pixel, since, that looks like more than 1 ;)

Re: 2c-Highground map mistakes and corrections

Posted: 17 Mar 2015, 07:44
by NoQ
But why is that still possible?
It doesn't work nearly as well against humans. Just put your lab at the entrance and it's an instant win against such strategy. You can even block the construction of this factory with a truck.

Re: 2c-Highground map mistakes and corrections

Posted: 17 Mar 2015, 17:38
by montetank
NoQ wrote:
But why is that still possible?
It doesn't work nearly as well against humans. Just put your lab at the entrance and it's an instant win against such strategy. You can even block the construction of this factory with a truck.
I know that, NoQ. I am speaking about all the players, playing against NullBot.

Re: 2c-Highground map mistakes and corrections

Posted: 17 Mar 2015, 19:29
by NoQ
If a certain chess AI always starts the game as white with 1.f3, 2.g4, it doesn't mean that black are overpowered or that the chess board needs fixing.

By the way i noticed some bugs in the updated highground map: gateways are completely misplaced (were not updated after terrain changes, i guess), and there's also a broken cliff texture near the north base. Will need another update.

Re: 2c-Highground map mistakes and corrections

Posted: 17 Mar 2015, 22:17
by montetank
haha-1-0 you win. I like, that you compares your Nullbot with a f3 g4 fool :lol2: