As I understand it, the question is not IF we fork, but WHEN we fork.
@zooko has already stated that he wants to change the parameters to incur less gas - I know it seems like he is currently saying no to fork of ASIC’s but he cannot change the parameters with out knocking the asics off the network. It is already too late for that.
Now, the question is this, is the network hashrate going to allow this change to happen in 9 months? If we go by more hashes = more say, then anyone without an asic cannot have a say. Are they really going to say “yes, make my machine obsolete”? would you? what protects the network after that?
There is very little incentive for a miner to use their ASIC hardware to protect the network in 8 months time when I know it will be kicked off in a months time.
Please look at the antminer empty block issue. Can you propose a method for bitcoin to fix this on their network? You really need to have a good plan for this because you will be facing this exact scenario when you try to fork to adjust the parameters.
This is what concerns me most. zcash is not a finished product (no crypto is) and by taking ASIC’s on this early, it severely limits your ability to react to issues with the PoW, or similar (segwit would break bitmains asics on bitcoin for example, so no segwit for the foreseeable future.) With CPUs, GPU’s and FPGA’s you still have the ability to adjust and fine tune the coin. You will never get this back with asics. See my previous post for more info.
I think it is worth noting that as you stated SHA256 is a lot simpler than equihash and had a lot more real world deployment. So which is more likely to have mistakes or need fine tuning? what about adding extra features?
I would really appreciate it if you could acknowledge that these are an issue.
and @daira thanks for using your voice in this to represent an alternate side.
Thanks.