It appears that the BitcoinCash chain is still on its original BCH main chain, and non-upgraded mines/exchanges/wallets can still transact there.
What then would happen if nine (or more) 1MB transactions were created by padding the comments so each transaction was just under 1MB. Assuming that these none-or-more transactions all had attractive Satoshi's per byte, is the following scenario possible? If not why not?
- Upgraded mines would try and fill a new block with nine (or more) 1MB transactions while non-upgraded nodes would try and fill a new block with exactly eight of them.
- If an upgraded mine won the next block, then it would create a physical 32MB consensus rule fork with an initial 9MB block off of the main chain, and we would be able to see this >8MB block in a BitcoinCash block explorer (at least for a while).
- Subsequent blocks mined by non-upgraded nodes would extend original 8MB non-upgraded main chain, whereas subsequent blocks mined by upgraded nodes would extend the upgraded now hard fork chain.
- Meanwhile exchanges/wallets which have not been upgraded (possibly some major players due to non-consensus seeking nature of this release) would operate only on the original main chain, while those that have upgraded would operate on the upgraded chain, assuming it is longer.
- Periodically the chain with the largest cumulative work would cause transactions on the other chain to be sent back back into the mempool, potentially allowing this process to repeat assuming that the new upgraded chain had less cumulative work and that the new upgraded chain again won the next block race and wrote a new 9MB hard fork block.
- Should this work, the overall effect of this could be tantamount to a temporary denial of service attack on the BitcoinCash network.