GossipSub - what is the expected behaviour for publishing a duplicate?

[CommoDor64]

Hey all,

Since I work on some analysis of GossipSub I try to experiment with publishing duplicates.

Technically, setting .duplicate_cache_time(std::time::Duration::from_secs(0)) on gossipsub::ConfigBuilder does set any element expiration to be immediate.

However, ,due to the Behaviour.publish(...)'s and TimeCache's design, it won't invalidate old entries in the cache until a new, non duplicate one enters. In other words, it will reject the immediate publish of the last message received (or any message with an identical key basically).

Is that the expect behaviour?

[CommoDor64]

NVM, I was too hasty and it seems to be answered before on #5278