I probably should a) set up a github repo b) place those patch links on top.. And we three should probably revisit the values before the upgrade!
That would be very helpful for others I'm sure. I'm interested in why your P2P_PORT is set to 23931 rather than the normal p2p port 13931 in your patched networks.py. Was that a custom value for your setup?
I read that the the "IDENTIFIER" and "PREFIX" can be generated for a new coin by taking 19 random numbers and converting them to hex, but that they should be consistent across all P2Pools. Assuming that's correct, I suggest we make your values for IDENTIFIER and PREFIX the 'official' values for any CGA p2pool for the sake of consistency across all the p2pools.
I'll try to set a github repo up today. I have to decide whether to branch from forrestv or rav3nPL - I think the second makes more sense because Rav3n is the king of altcoin p2pool, but so far I went with forrestv.
There are two P2P_PORT variables (4 if you count testnet), the client's P2P_PORT which is defined in p2pool/bitcoin/networks.py, and P2Pool's own P2P_PORT which is defined in p2pool/networks.py. It would be impractical to use the same port numbers there - you would get a conflict unless you run the client on a different interface. So I just kept with the 'tradition' of using the same RPC & P2P_PORT as the client, but raising one number - like with bitcoin, client's RPC/P2P ports are 8332/8333, P2Pool's ports are 9332/9333.
The IDENTIFIER & PREFIX variables are indeed random (although 16 characters IICCC*), and serve as a fingerprint. Keeping them consistent is also indeed the best way to keep a strong P2Pool network. One thing though - if we're going to change values, whether it be block period or spread or whatever, it would be a good idea to generate a new fingerprint to avoid internal conflicts.
*If I Can Count Correctly