Merged mine coins are still different coins, Protocol versions can always be changed or the merged mining code can just use the appropriate version number when communicating with the merged coin.
The difficulty floor is also not an issue, there are going to be times when the merged coin might have a higher difficulty then the parent coin, merged mining simply stops until the merged coin difficulty drops below the parent coin.
I could not leave the difficulty floor where it was at. The floor in Blakecoin was fine when Blakecoin was CPU only, but there is absolutely no way I could start the new UFC blockchain that low with GPU/FPGA miners.
During testing, just one of my rigs would totally dominate the network and generate blocks faster then the network could process them. Could you imagine the cries of "instamine" with the starting difficulty Blakecoin had? The floor I have set now will allow the network to go down to 3-4 GH and still maintain the 1 minute target spacing.
In short, nothing I have changed should effect adding merged mining in the future.
"In short, nothing I have changed should effect adding merged mining in the future."
I have been working with the pool quite a bit I don't think this is true! it has created the issue that a merged coin will be based on one of the forks not both, you would need to copy some of the function in the pool software to be able to process both chains e.g the address and other bits mentioned (pool only assumes one type of address not a blake-256 for one and sha-256 for another, things like this cannot be changed after without messy hacks!)
"Merged mine coins are still different coins, Protocol versions can always be changed or the merged mining code can just use the appropriate version number when communicating with the merged coin."
I have not seen this do you have example code? did you read the merge mine proxy for eloipool?
https://gitorious.org/~Luke-Jr/bitcoin/luke-jr-bitcoin/blobs/namecoin_mmm/contrib/merged-mine-proxy"network to go down to 3-4 GH and still maintain the 1 minute target spacing"
the law of large numbers controls target spacing as it is an average of the past random events (block find) with the diff set to create the perfect x time block, you can see an example of this in the all time average of my pool 100.70% so it should have better averages at any diff due to Blake-256
"Could you imagine the cries of "instamine""
yeah I can hear the cries now: static const int64 nSuperBlockCoin = 8000000 * COIN;
code speaks for itself:
https://github.com/scificryptocoin/Federation-Credits/blob/master/src/main.cpp#L1083"merged coin might have a higher difficulty then the parent coin"
this would only happen if you had a slower block time but you state 1 minute? so it should not have a higher diff?
so much for the concept of united/merged compute/mining at this rate Bitcoin will always have the upper hand, Scrypt based coins prove this
I am sure many of you have seen this motto "United we stand, divided we fall"
if anyone has a plan for a Blake based coins please contact me before hand and spend just a little longer testing the wallet and we can have merged/standalone coin it does
NOT have to be merged only with eloipool merge mine proxy, address and some other small bits need to remain else you need to do some re-writes as the merge mine proxy re-uses the pool functions which have been modified for Blakecoin and as explained above, this hybrid does not use same functions thus it is incompatible without quite a few changes