Jeff Garzik is developing a new piece of software, intended to implement the New York Agreement, which can be seen here:
The software will flag bit 4, but not bit 1, which means SegWit will not activate and existing nodes will not allow witness data on the Bitcoin p2p network already in use. This means a new p2p network may need to be set up, which could take many months of development and testing, and means the network will be very complex, with three p2p networks.
Flagging both bit 4 AND bit 1, such that the software is compatible with both SegWit and the NY Agreement can be done. It has even been implemented in BIP91. There is no clear reason for rejecting BIP91, and doing so appears destructive and potentially malicious.
When questions Jeff stated:
So the real answer is that we want to be maximally compatible with segwit within the bounds of the charter – a safe network upgrade to segwit-AND-2m. Post-hashpower-activation segwit2x will be the only segwit, for all intents and purposes, so deviations pre-activation would simply be legacy code post-.
I am not sure what this response means. It appears as if the software needs to do something unnecessarily stupid to be compliant with the "charter". Is this really what NY Agreement participants signed up for?
Please can we end these stupid games and do what is best for Bitcoin. It seems the NY Agreement has intentions other than SegWit and a hardfork later on, but instead some participants seem intent on causing delays, complexity or trouble.