Changed to nTimeBlock

On second thought, why do we need to make a hard fork upgrade for this feature, as the purpose of IsSuperMajority() is to allow miners/pools to migrate at their own schedule?

I can understand it's a problem on testnet, our testnet has no mining majority as likely no one is mining on testnet most of times, so if your new node mines 100 blocks your node would falsely think 'the majority' has upgraded so it starts rejecting other miners' blocks

It is not hard fork, it is a soft fork.

The v06 activation time prevents the softfork from being activated prematurely, which happens on testnet and therefore testnet fails to sync.
It happens on testnet because there are a lot of v2 blocks already in the chain.

Just try to sync the testnet with the client without this fix, you'll see.

IsSuperMajority() check both PoW and PoS blocks, it's not about miners only on peercoin.

Yes, I meant both miners and minters. If I understand correctly, is it not only limited to testnet? On the testnet, isn't it only a problem as there is no consistent minting/mining body? Or else how did you reach 75 blocks of 100 to trigger the enforcement of version 2 blocks? What was the concern such that mainnet must use a delayed turn-on of this feature?

The soft fork get's activated on testnet because someone experimented with version 2 blocks on it.
Indeed the switchtime is not needed on mainnet, but it is much cleaner and more readable to introduce a V06 switchtime compared to adding if(testnet && islaterthan())

What alternative do you suggest?

I am thinking maybe we should choose only one of the two:
a) bitcoin's IsSuperMajority()
pros: no mandatory upgrade deadline
cons: users not sure when the upgrade deadline is, especially not nice for miners
b) pre announce mandatory upgrade deadline like peercoin's v0.5, v0.4
pros: clear expectation for users, easier to understand behavior for debugging
cons: requires every node to upgrade before deadline
I think maybe b) should be our preferred approach, of help with our testing on testnet and also easier to diagnose if something happens. bitcoin later versions have quite a few of these 'soft' upgrades, so it's also a consideration down the road. this means disabling the calls to IsSuperMajority(), treating the bitcoin soft upgrades as mandatory protocol upgrades.
For this particular upgrade, all the nodes that mint blocks (pos and pow) are required to upgrade, unlike in bitcoin's case it's only the mining nodes.
What do you think?

I don't know if you see that this pull request currently implements a) using bitcoin's IsSuperMajority()
It just makes sure that it can't happen before V06 switchtime which is currently even disabled.

I believe we should go for a) if we won't make any protocol changes anymore, as it would allow old clients to still validate the chain, they will just fail to mint.

I agree that if we are going to hard fork for other reasons, we should hard fork this change too.

Because all of this is not yet clear, I suggest to just merge this pull request. Currently the soft fork is disabled anyway (see code snippet below). At least we'll have a working client on testnet, I've been running and minting with this code on testnet for a few weeks now.

bool IsProtocolV06(unsigned int nTimeBlock)
 {
     // TODO decide on upgrade timing
     return false;
 }