[Routine]
No details yet.
[PlayerMover]
Problem:
Player movement is a very important aspect of Exilebuddy and botting in general, but the current PlayerMover system was not designed to handle the current state of the game. As time has passed, the meta of the game has evolved into clear speed, which requires good player movement logic which does not currently exist outside of community plugins.
Community plugins for this topic are rough, to say the least, because the underlying system for handling player movement was not designed for the current state of the game. The current system is a legacy system, so it's been on the todo list to update for some time now.
Solution:
The PlayerMover system will be updated to a system similar to what Routines use, bringing the core design of player movers into the modern age. This should help simplify the issue a lot, and provide users with more usability with minimal change. The specific details of these changed were made here:
https://www.thebuddyforum.com/threads/notice-upcoming-playermover-breaking-changes.418088/
The changes exist in the current builds, but the whole system is not activated yet, to give people time to look at the system and update their code or offer suggestions. We do intend to have our own skill based player mover in the future using this system, but community plugins for it are always welcome of course.
[Input]
Problem:
One of the long time issues with the way the API performs input actions is that it warps the mouse to the target destination, and does not support "moving" the mouse in a human-like way. This has created known side effects due to the way this game works, but they have been manageable for the most part to where the system didn't need to change.
Solution:
A new MouseManager class (name might change in the future) has been added and is currently being worked on. This class actually uses the same system as the legacy PlayerMover system does (this too might change in the future), to allow devs to implement code to override the mouse movement right before an API mouse move is performed to the target destination.
With this in place, optimal improvements can be made by devs to Exilebuddy to make it appear more human-like from an input standpoint, and the side-effects of the mouse warping issue should be resolvable finally. There's still a lot of updates to make this a reality, but the plan is to have the system in place and usable before 3.3 arrives.
More specifics and examples of how this is useful for EB will come in the future.
[Pathfinding]
Problem:
The current pathfinding library is from 2012 - 2013, and is in need of updates. While the current implementation has been "good enough" for a while, it's time to finally update the library and our Pathfinding API used in Exilebuddy.
Solution:
The latest version of the pathfinding library will be updated to first. After Alpha testing is done to make sure nothing obvious breaks leading into 3.3, we'll update the pathfinding wrapper the API currently uses. This will result in a lot of breaking changes, but should hopefully cleanup a really messy system.
In the future, we hope to make more improvements to bot logic making use of the new pathfinding library. The pathfinding breaking changes thread lists several known problems and how we'd like to solve them, but there's also issues that cannot be solved easily. As a result, we hope to get most of the most critical issues with pathfinding taken care of for 3.3, and give ourselves room to improve with the logic in the future.
[Atlas]
Problem:
Atlas support has been missing since it's inception. We have had requests for supporting this API but there was simply not enough time to get it worked out.
Solution:
Currently in Alpha and in the next Beta/Release, support for the Atlas GUI has been added. Some preliminary testing shows it's working as expected, and it shouldn't be too hard to maintain going forward. We hope this addition will help improve the dynamics of Exilebuddy when it comes to the end game grind!
