New catch version is pretty long

[raskyer]

Expected Behavior

Can catch pokemon and find apropriate path

Actual Behavior

Makes long minutes before actually seeing a pokemon

Why yesterday behaviour has change ?

[fredrik-hellmangroup]

what commit are you running? i dont see any issues on b8606bc

[2016-07-25 13:36:24] [#] A Wild Pidgey appeared! [CP 125] [Potential 0.42]
[2016-07-25 13:36:24] [#] IV [Stamina/Attack/Defense] = [6/2/11]
[2016-07-25 13:36:26] [x] Using Pokeball (chance: 53.30%)... (87 left!)
[2016-07-25 13:36:27] [x] Captured Pidgey! [CP 125] [6/2/11]
[2016-07-25 13:36:27] [#] Pidgey has been exchanged for candy!

[tstumm]

There is an issue regarding SpiralNavigator. @douglascamata is already working on a fix.

[raskyer]

I'm on 'dev' to be as close as possible on the latest version and try to give some work on it.

@tstumm Thanks ! Waiting for this fix :)

[douglascamata]

I'm having trouble with the fix. Any help is appreciated. Please setup the maps integration to have a better picture of how (bad) it's walking. lol

[raskyer]

@douglascamata I will try to see that tonight (I'm at work for now). The former method wasn't enough great ? Because yesterday it was actually pretty efficient.

[douglascamata]

@LeaklessGfy we're refactoring towards implementing a Behaviour Tree and this broke the SpiralNavigator. It's not calculating next points to walk to correctly, I guess. :/

So, tips for your investigation:

On a 1st tick this happens:

1. SpiralNavigator calculate where it wants to go (let's call it point A) from the current position (let's call it point P) .
2. It uses StepWalker to take a small step from P to A

On 2nd tick:

1. SpiralNavigator check the bot's current position is equal to A (its desired destination).
2. (a) If not at A yet, use StepWalker again to take a small step from current position to A.
3. (b) If current position is equal to A: calculate next destination (let's say point B) and walk to it

This way we can implement different navigators, like a SquareFarmNavigator, for example. Or CityMoveNavigator to walk from one city to another without getting insta softbanned.

[raskyer]

@douglascamata Alright thanks for this informations and can you tell me more about the problem you encounter ? What's actually happening with that function on map ? Google map direction make the process buggy ? Another thing ?

[douglascamata]

@LeaklessGfy basically the bot is walking like a drunk cockroach. I think next step is not being calculated correctly by SpiralNavigator. Probably needs to plot all the points calculated by SpiralNavigator in a map with, lets say, red color and StepWalker steps in blue color and compare to help understanding.

[raskyer]

Alright the big problem here is that the master branch is actually at the same point of dev ... I think that's a terrible thing because here we are stuck until we don't find the solution of this problem and I can't get a eye on it for now ... On slack we should warn every dev on this issue.