kano, it's understandable for ck to not use ckdb as he's mentioned that the db stuff is beyond his remit. As a result he has a dependency on you to test that commits don't introduce retroactive bugs in ckdb. You're really the best person for that as you run it in production and are the only person doing so openly. If you're using an old version of it (August 2016), the private fork happened then, not now.
ck's been recently making changes to ckpool for ckdb, for the Bitmain guy, coz he paid him to do them, that brake the ckdb payouts.I hardly see "beyond his remit" in that.
I am not the only one who runs ckdb in production, others do.
It was 2 other people, BitClub and RogerVer, who had ckdb startup crashes when they updated with ckpool changes that were for ckpool - but they were changes that affected ckdb also.
Normally if you change a library function that requires new rules about calling it, you'd change all code that calls it, not ignore the ckdb code that calls it and was broken by the change.
As for when the fork happened - who cares.
He has 2 forks of ckpool.
I created one and he commented positively about it when I said I created it.
Then the other day he got pissed at me responding to his jibe at me in similar fashion, then locked me out of everything saying it was coz of my fork. Oddly that included locking me out of cgminer and taking away all my privs in the IRC channels.
Nothing to do with code or forks, he was just throwing a tantrum coz of what I replied pretty much similar to what he said to me.
You feel the need to lecture me for -ck
I'll respond.
P.S. this is over, finished end of story. Not sure why you felt the need to bring it up again.
As is CKDB as per his words after he locked me out:
"< conman> leaving all the others depending on ckdb in the lurch yay \o/"
... The work you both do is as inter-dependent as it is important.
He certainly doesn't agree with you.Quote
Re: the private CDF calcs: I can digg that it very much classifies as secret-sauce for a pool. If a pool is a government, defense materiel would definitely include custom, unshared BWH-protection and/or luck-based pool operator actions. Warnings that lead to bans and eventual non-payment is serious shit and affects your reputation as a solid pool operator. So you have to protect that.
Nothing to do with private CDF calcs.Some of the current testing is there in the public git and the database and code changes for the stats are all there also.
There's ways a miner can make it very hard to detect their withholding.
This new testing is doing something very different to detect that, and I will be able to check things retroactively with some effort later - since I always keep all data.
. Let's hope that when your second R4 arrives it brings us back some green luck! 
and guaranteed 
... in the past month is jumped from 310.... to 392... and bitmain start shipping the next batch s9 on feb 20 , that would give much more higher difficulty by the end of February .