willowtreeapps / wombats-api

Wombats API
https://github.com/willowtreeapps/wombats-documentation
MIT License
10 stars 4 forks source link

Fog of war walls are displayed as fog #52

Closed oconn closed 8 years ago

oconn commented 8 years ago

This bug occurs when a bot is placed next to a wall at a 45° angle. The wall at 45° is returned as fog.

? ? ? ? ? _ + _ _ _
? ? ? @ @ _ _ _ _ ?
_ ? ? ? _ _ _ + ? ?
+ _ ? ? _ _ ? ? ? ?
_ _ _ _ B @ ? ? ? ?
_ _ + _ _ _ ? ? ? ?
_ _ _ _ _ _ @ ? ? ?
_ + _ _ - _ ? ? ? ?
_ _ _ _ + _ _ ? ? ?
_ _ + _ _ _ _ _ ? ?
_ _ _ _ + _ _ _ ? ?

cell (3, 3) in this arena is a block.

oconn commented 8 years ago

@thoughtmanifest @dehli

When applying the occlusion logic to the user partial I was able to gain a little more context around this bug.

image

while the bot is not depicted in this image, it is located at position (10,10). (note row ten is printed a half step to the left)

It appears that (10,9) is casting fog on all the walls running from (0,9) to (9,9), or is interpreting them as walls an is converting them to fog. Same applies to the wall running from (11,10) to (11,20)

thoughtmanifest commented 8 years ago

And what does the non-occluded arena look like?

On Wed, Jun 29, 2016 at 11:35 AM, oconn notifications@github.com wrote:

When applying the occlusion logic to the user partial I was able to gain a little more context around this bug.

[image: image] https://cloud.githubusercontent.com/assets/4649439/16458411/f4dac13e-3dec-11e6-80df-88af28c4bc44.png

while the bot is not depicted in this image, it is located at position (10,10). (note row ten is printed a half step to the left)

It appears that (10,9) is casting fog on all the walls running from (0,9) to (9,9), or is interpreting them as walls an is converting them to fog. Same applies to the wall running from (11,10) to (11,20)

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/willowtreeapps/battlebots/issues/52#issuecomment-229395195, or mute the thread https://github.com/notifications/unsubscribe/ABAGg3GOD0Lq9uCTQidTJ9WGPbZ-NYElks5qQpC8gaJpZM4I_99X .

oconn commented 8 years ago

This is not the exact arena, however same concept

image

(10,10) is the upper right corner of this arena

thoughtmanifest commented 8 years ago

I should have been more specific, I was looking for the non-occluded version of the player arena in order to see what were actually walls and what weren't.

Also, could this have been affected by the use of differing interpretation of the x and y axes?

oconn commented 8 years ago

So I fixed all the x y interpretations in the PR that I have up now. Each row represents a y axes.

image

image

thoughtmanifest commented 8 years ago

I'm glad to see those were actually walls, I was worried the 9 and 11 axes weren't walls. I'll see what I can do about that this evening unless @dehli has some thoughts.

re: Arena axes, so you're able to use (get-in arena [x y]) to return the x,y coordinate? or is it (get-in arena [y x])?

oconn commented 8 years ago
(defn get-item
  "gets an item based off of given coords"
  [[x y] arena]
  (get-in arena [y x]))

I'd use this method under battlebots.arena.utils. We are now testing all of those funcitons