Difficulty Rising


#177

It is the size of the spike that concerns me. 5GH down to 0.5 then back up again? What is up with that?? Are we seeing someone getting ready to try an attack on the ZCash network?? I remember seeing things similar to this with the other equihash coins before hand as well (testing their capability?). @Zooko @Daira you guys keeping a watchful eye on this?


#178

The up-and-down of the EFFECTIVE (not actual) hashrate and the difficulty is what maintains the two-and-a-half-minute block time


#179

I understand that, however current hashrate atm is 1.79GH. If we are seeing spikes up to 5GB, then it worth paying attention to as that is by far more than enough to cause trouble on the network if it is a concerted effort of a single person/group. Scenario’s like this, even in the short term can make it possible for double spends, and other kinds of shenanigans.


#180

might be miners shutting down workers before difficulty retargets. large miners try to game difficulty by shutting down before a retarget, hoping for lower difficulty, then fire workers back up with larger hashrate.


#181

that would explain the drop down to .5GH, but it doesn’t explain where and extra 3GH of hashrate suddenly came from. I assume the severe drop was from all the auto-switching miners jumping to another coin chain.


#182

Actual hashrate and effective hashrate
It’s not actually 5Gh, its 1.6 getting lucky <2.5min
So the difficulty goes up
Then its not actually 0.5, its 1.6 with less luck >2.5min
So the difficulty goes down
Repeat
Its how it works


#183

I see what you are saying now. The numbers aren’t the actual hashrate, it’s what the hashrate “looks” like based on block luck at that moment in time. Now I’m following you.


#184

If i understand correctly, difficulty adjusts quickly so even a very large change in hashrate would resolve within 20 blocks.

Not much opportunity for miners to game it, doubt its even worth trying.


#185

Agreed, unless it was “actual” hashrate, then there is the 51% risk, which is what I originally worrying about until it was made clear this was a representation of what hashrate “looked like” based on block luck.


#186

Bitmain is using Antpool, right? They are back on top of the pool list…what are does guys doing?


#187

If i remember right (pretty sure though) Bitmain uses always a secret/private pool within antpool.
Means their hashrate/power is not directly visible/shown under Antpool but under an external/seperate
mining pool ZEC wallet adress.


#188

not saying i’m correct on this one, but i’m sure that doesn’t stop them from trying.


#189

ZCash difficulty gets adjusted every block…


#190

Very true, however there’s dampening in the diff calc so a huge hash change takes a while to normalise. Way better than BTC.


#191

Yes, not dampening would make it unstable. I’m very happy overall with how well the difficulty adjustment works, although it does have some problems responding promptly to large drops in difficulty (as currently seen on testnet).


#192

I spoke with Bitfly (the owners of the zChain explorer and Flypool) in the private chat and they clarified why the pool hashrate distribution chart has a Founders Reward address on it.

It’s not due to a miscalculation but it’s how a pool puts the Founders Reward address first and thier personal address second, while all the rest of the pools are the opposite. So according to them that slice of the pie is a pools hashpower (Antpool in this case) which is represented by the Founders Reward address. That explains why it’s not exactly 20% and also shows that the pool distribution is in fact better than we had assumed by excluding that slice of the pie.

They claim to have fixed the underlying bug: https://zcha.in/blocks/00000000006a542f233cd87bb722f4c79e71fc21dea6c2343afd740dc38a265c but it still seems to persist


#193

nice…difficulty raising and price bleeding, what a time to be a Zcashier…


#194

I love it pump it to 100M :smiley:


#195

aaand difficulit -20M right now :exploding_head:


#196

That is one huge spike and that hash had to go somewhere. What do you think is going on? Are ASICS manufacturers just testing hardware before selling them or what ?