Core/Movement: fix the issues regarding MovePoint not handling correc…#17028
Core/Movement: fix the issues regarding MovePoint not handling correc…#17028chaodhib wants to merge 1 commit into
Conversation
…tly rooted units.
|
ermmm, right xd I've already done all the work here, lets not duplicate efforts xd |
| void MotionMaster::ResumeMovement() | ||
| { | ||
| MovementExpired(); | ||
| } No newline at end of file |
The idea is to have a PR with only cleanup changes, maybe little changes in small PRs (like "Renamed _something() to Something() ") to allow a reviewer spot possible issues |
|
Any reason to don't merge this? |
|
|
mutate is not the way to go. |
|
@ccrs where is your pr on this? |
|
This PR introduces a critical bug with mob pathing. Repro:
|
|
due to the nice logic of duplicating and separate work on 52343241 PRs its simply not updated, but got it working on everything |
|
@ccrs got a branch I can play with? Going through your open PRs and reviewing atm. |
|
The PR has some holes like the scenario described by Treeston and is not ready to merged. However this could be solved by adding another layer in the MoveGen stack (remove the size limit of the stack which is now set to 3). I don't think the idea behind this PR should be dismissed since it's simpler than ccrs's solution because I'm reusing existing components. |
|
@chaodhib MoveGen should probably be rewritten to use std::priority_queue with a lot of possible slots |
|
Conflicting files |
|
wow, part of my issues in razorscale is solved here o.o |
|
@chaodhib in #19828 i'm having issues with razorscale making "teleport" instead move after add UNIT_STATE_ROOT and remove UNIT_STATE_ROOT. 2 of my issues looks like fixed by this pr |
|
Ok, I'll investigate as soon as I can. |
|
@Keader totally not related, you simply erase slot_active with a movepoint.... to put it simple MovePoint erases MoveChase because they share the same slot. 1 : 1 script developer manual. no issue or bug on motionmaster. |
|
Agree that an api allowing to resume chase after movepoint would make scripting more intuitive. |
|
@ccrs where to see your mentioned PR? |
|
@Faq : his TrinityCore PRs are listed here: https://github.com/TrinityCore/TrinityCore/pulls/ccrs |
|
Conflicting files |
|
Superseded by #21888 |
This fix issues regarding Motionmaster::MovePoint() not handling correctly rooted units:
The PR is not ready to get merged as is. This is a draft. It needs some polish but the fix seems to be working.
Testing done:
I tried to make the least amount of changes so that the PR is as readable as possible. I think however that the whole movement system needs some serious clean up and this means unreadable and massive PRs in the future I'm afraid.