Take a look at testnet today. We're in the middle of a brutal test and it's looking very good!
http://testnet.geekhash.org/
This phase of the test started at around 1700 - you can see the drop in the red 'difficulty' line as the test was reset. The approx 2MH/s was my and Blaksmith's hashing. TheQueen74 paid us a visit from the City of Light and kicked things up a bit. The block times look good overall in spite of some ~30 minute blocks on the 'down hill' side of things.
The days of long CAT blocks and slow confirms are almost over.
http://testnet.geekhash.org/
This phase of the test started at around 1700 - you can see the drop in the red 'difficulty' line as the test was reset. The approx 2MH/s was my and Blaksmith's hashing. TheQueen74 paid us a visit from the City of Light and kicked things up a bit. The block times look good overall in spite of some ~30 minute blocks on the 'down hill' side of things.
The days of long CAT blocks and slow confirms are almost over.
Looking good on the testnet!
Is that still using the 12% lower limit? A good PID should obviate the need for the lower limit, it just makes for some very very long blocktimes after an attack. If it's been a couple hours since the last block, let the diff drop up to 25%.