Replicator bug

[Terkala]

What version did the issue occur in?

No response

Description

This bug appeared on Funky, but we have the most recent version of Replicators so it should be similar enough. Round replay is space-wizards#15309 Cirno :

the hole got destroyed by blob, a replicator was t1 and trying to upgrade while in harm mode, ended up spawning 100 replicators

Reproduction

1. Destroy Replicator Hole
2. Have replicator T1 and mid-upgrade and have harm mode on
3. Clone 100+ times.

Additional Context

No response

[fiamys]

Round replay for the curious: https://replays.iterator.systems/replay/funky/LRP1/15309

Replicated on an imp localhost by simply spawning a repli, and manually deleting the hole. No need to be in harm mode.

We dont have blob on imp. And I believe the hole is immune to rods and such, which means the nest never gets deleted outright through game mechanics, only destroyed by force by players. I assume the deletion that the blob performs skips the things that normally happen on hive destruction. Replicators on imp normally downgrade to tier 1 when the hive gets destroyed, which hasnt happened in the replay either.

Theory would need to be tested by trying to replicate with one person destroying the hole with a knife while another tries to duplicate itself, but I only have myself, so I cant really do both at the same time

[fiamys]

Okay, I tested rodding and it is not immune to that. However, destruction by rod did give me the option to create a new hole, and upgraded the queen to tier 3/downgraded the other replis to tier 1 appropriately, which the manual deletion did not, so I'm still going to assume blob shenanigans for now

Edit: Actually, rods just do a shit ton of blunt, dont they? I guess that would still count as destruction by force, then. Which would mean the only way for this to happen without admin intervention on imp is via tesla or singulo (Both tested and successfully replicated)

[Terkala]

Odd, blob just does blunt damage, it doesn't do any special deletion logic

[Terkala]

But anyway, if you've confirmed this doesn't happen on Imp currently, that's good. We'll be moving off of our current version of Replicator and losing Blob anyway with our Forky move, so this isn't critical to resolve that special case.

[fiamys]

If the Blob doesnt have deletion logic, I dont understand why the hole is destroyed in the replay. It just... vanishes the first time around.
This has apparently happened on imp before. I can't see a bugfix merged related to it. Would probably be nice to keep open for reference later down the line if it becomes relevant again?