[2.0.21] Set filter on Inserters and Asteroid Collectors behave differently.

[IsaacOscar]

Consider the following setup:

Screenshot 2024-11-26 102323.png (574.27 KiB) Viewed 399 times

Screenshot 2024-11-26 102428.png (751.43 KiB) Viewed 399 times

The constant combinator is set to output 1 "Oxide Asteroid Chunk".
The power pole reports 1 "Oxide Asteroid Chunk" and 1 "Metallic Asteroid Chunk".

Despite the metallic asteroid chunk = 1 signal, the filter of the collector is only set to oxide chunks (because the metallic asteroid chunk is from the "read contents" of the asteroid collect, and so is ignored). The asteroid collector appears to be working as intended according to boskid: viewtopic.php?p=641115#p641115, and I agree it's behaviour makes sense.

However, inserts do not behave the same:

Screenshot 2024-11-26 102703.png (98.97 KiB) Viewed 399 times

Screenshot 2024-11-26 102709.png (307.85 KiB) Viewed 399 times

The constant combinator is outputting 1 "Pipe".
The power pole reports 1 Pipe and 1 Iron Stick.

However, the filter on the inserter is set to both Pipe and Iron Stick. Indicating that it is not subtracting it's own "Read Hand Contents" signal.


The bug I'm reporting here is the inconsistency, however I would prefer the asteroid collector behaviour.

[Soul-Burn]

There are more cases such as these inconsistencies in this reddit post.

Assemblers see their own signals (but not for "set recipe").

Collectors don't see their own signals (as expected).

Inserters see their own signals (as you mentioned).

Requester chests see their own signals.

[Rseding91]

Thanks for the report however we never said every entity used the exact same behavior for how bits of them work. This is feature request not a bug.

[IsaacOscar]

Thanks for the report however we never said every entity used the exact same behavior for how bits of them work. This is feature request not a bug.

Fair enough, however it seems rather arbitrarily inconsistent, and is confusing and undocumented.