Chaco
Level: Smitemaster
Rank Points: 3626
Registered: 10-06-2005
IP: Logged

File: G GGG 1N1E weird gentryii brain pathfinding behavior.demo (4.1 KB)
Downloaded 42 times.
License: Public Domain

While playing Dying Flutchman's "Gentryiification" hold in Architecture, I ran into some strange pathmapping behavior where a Gentryii would attempt to pathfind through one of its own diagonal chain links. Or, at least that's the only explanation I can come up with for the behavior in the attached demo for Gentryiification: GGG: 1N1E.

However, as Caravel Chat noted later, building a fresh test room and placing down a Gentryii immediately in the given position, results in the Gentryii moving Northwest and around the blockade like normal.

So, I think there's some kind of failure to properly update brain pathmaps when a chain is being dragged around, or something.

EDIT: For completeness I should also say that this occurs in both the Mac build 6512, and the most recent Windows build, so it includes this earlier bugfix with Gentryii chains and direct-moving enemies.

____________________________
Quick links to my stuff (in case you forgot where it was):

---

[Last edited by Chaco at 10-03-2018 02:05 AM]

Dying Flutchman
Level: Smiter

Rank Points: 406
Registered: 01-27-2017
IP: Logged

Hi Chaco,

You might indeed have found a bug. During playtesting this hold, I sometimes had strange things happen. However, gentryii are complicated, so I've bassically ignored these occurances. Once or twice I tried to reproduce, but that did not work. I just blamed it on an oversight on my part and did not bother to analyse the movement.

Your statement that it might have to do with the updating of the pathmap seems to make sense. That's consistent with the difficulty of reproducing...

Regarding the move the gentryii should make in this case: NW should be correct. Gentryii chains are simply brain-visible objects. A gentryii will, however, treat a gentryii head as just another monster that it will happily pathmap through.

All this often does not really matter, because most rooms you encounter don't depend on predicting precise gentryii movement. And this is a *good* thing

____________________________
Autocorrect is not my friend. Apologies for the typos.

Xindaris
Level: Smitemaster

Rank Points: 1531
Registered: 06-13-2015
IP: Logged

It seems to me like DROD just really hates updating the brain pathmap to account for things that move around...

____________________________
109th Skywatcher

Here are some links to Things!

skell
Level: Legendary Smitemaster

Rank Points: 3734
Registered: 12-28-2004
IP: Logged

File: GentryiiBug.gif (256.4 KB)
Downloaded 181 times.
License: Public Domain

Here is a visualization showing the problem. The white squares on tiles show directions which are blocked from pathmapping on that tile, and the cyan hilights show which tiles were recalculated for pathmap last frame. You can actually see there was another problematic case (though most likely inconsequential) near the head.

____________________________
My website | Facebook | Twitter

skell
Level: Legendary Smitemaster

Rank Points: 3734
Registered: 12-28-2004
IP: Logged

And fixed, PR. There is real risk it will break some demos (hopefully not rooms) but this deserves to be fixed.

____________________________
My website | Facebook | Twitter

mrimer
Level: Legendary Smitemaster

Rank Points: 5056
Registered: 02-04-2003
IP: Logged

Thank you for digging into and fixing this issue. I think you got it!

Love the pathmapping debug display you've implemented!

Wondering if that should be a releasable thing that can have the options stepped through by using a function key, or via a console command, or somesuch.

____________________________
Gandalf? Yes... That's what they used to call me.
Gandalf the Grey. That was my name.
I am Gandalf the White.
And I come back to you now at the turn of the tide.

kieranmillar
Level: Smitemaster
Rank Points: 2670
Registered: 07-11-2014
IP: Logged

Bumping this to ask someone for help verifying this, I am very confused as to what the cause of the bug is here and how to know if this issue is fixed. It is not clear to me from the cool gif what the actual problem is. Is it that it is moving to the left despite not having any viable brain path to the player at all?

I have not played the hold in question and so cannot access the rooms for the demos in the OP in order to verfiy this.

skell
Level: Legendary Smitemaster

Rank Points: 3734
Registered: 12-28-2004
IP: Logged

File: IsItClearerNow.png (89.6 KB)
Downloaded 105 times.
License: Public Domain

Does it help?

____________________________
My website | Facebook | Twitter

kieranmillar
Level: Smitemaster
Rank Points: 2670
Registered: 07-11-2014
IP: Logged

File: BUG REPORT Brained gentryii pulling.hold (938 bytes)
Downloaded 39 times.
License: Public Domain

Thanks skell, I think I get it now.

As far as I can tell, in 5.1.1.alpha.2020-10-18, gentryii brain pathmapping around chains seems to work OK.

There is another issue I've found with brained gentryii though, regarding pulling chains tight. They don't seem to want to do it.

Check the attached example hold. Just wait a while near the brain until the gentryii stops moving, then kill the brain. Note that while brained he wouldn't tighten the chain.

skell
Level: Legendary Smitemaster

Rank Points: 3734
Registered: 12-28-2004
IP: Logged

File: Mr T Ate My Socks.png (89.1 KB)
Downloaded 90 times.
License: Public Domain

kieranmillar wrote:
There is another issue I've found with brained gentryii though, regarding pulling chains tight. They don't seem to want to do it.

Check the attached example hold. Just wait a while near the brain until the gentryii stops moving, then kill the brain. Note that while brained he wouldn't tighten the chain.

That works as intended. For reference, here is the situation:



If you stand where I stand Gentryii keep alternating moving south and north, because Brained monsters want to move as long as it's possible to move. You can see the same with swarms of roaches trying to fit in a corridor, they'll trash around semi-randomly at times. If a brained monster can't move closer to the player but can move to a tile that's the same distance, they'll do jut that.

If you stand on A and face SE you'll block the gentryii movement and it'll tighten the chain.

Similarly if you stand on any of the B tiles there will only ever be one tile that's closest to the player and it'll straighten it again.

Thanks for staying on your toes!

____________________________
My website | Facebook | Twitter

---

[Last edited by skell at 10-18-2020 08:11 PM]