Quote from: amaclin on October 27, 2016, 08:28:59 PM

Quote from: AgentofCoin on October 27, 2016, 07:21:24 PM

What would be a legitimate reason to push non-confirmed transactions over and over?

Do we need a reason to broadcast valid non-confirmed transaction to a peer?

Yes, so all peers will propagate that valid unconfirmed tx.
But why push a valid unconfirmed tx that shall not be accepted into any valid block ever?

Quote from: amaclin on October 27, 2016, 08:28:59 PM

Quote from: AgentofCoin on October 27, 2016, 07:21:24 PM

Isn't re-boradcasting it over and over and keeping it in the mempool just adding to the
problem as to

I do not know.

Isn't that a memory usage attack vector?

Quote from: amaclin on October 27, 2016, 08:28:59 PM

Quote from: AgentofCoin on October 27, 2016, 07:21:24 PM

It is understandable if your tx is good and after a few days you repush it when a bottle
neck is over, but nodes pushing for weeks or months seems like misbehavior.

No, it is not understandable.
You have a right to locate all such nodes and punish their owners.
You also have a right to keep forever all tx hashes for the transactions you have deleted and ban
the peers which re-send these transactions. Grin

Here is the question then:
Can Bitcoin attacks and attack vectors be a malicious act or a misbehavior in your eyes?
Or since it is doable or executable, it is always a justly act and there can be no injurious acts.