I'm more convinced than ever that BTC is basically dead...
...
4) Centralization problems haven't been solved. In the last two years, two major 51% scares have happened, both involving ghash.io. Each time the alarm bells ring, new ideas are offered, and absolutely nothing changes.
You are right on all counts, but on any given day DRK's hash is usually even more centralised than BTC's.
Evan and team need to
a) Admit the fucking problem
and
b) Fix it by shifting blockchain security to Masternodes. Most of the code and all of the infrastructure already exists. Blockchain security can and should be practically unbreakable, the same as Darksend and InstantX,
not left in the hands of two or three pool ops.
Go talk to anyone with serious money about trusting big chunks of it to a mightily 'decentralised'
3 shitty pool servers, all run as cheaply as possible by some noddy from the interwebs, see what reaction you get. Then have the same conversation again but with
3000 servers, n-of-m subsetting, 1/(10^20) attack success probability... your charts might suddenly start to look a bit more exiting than less than a fiftieth of the price of crappy old Bitcoin.
And what about a single individual could own half of the masternodes? Financially it's not making much sense but this COULD happen. You're one who always talks in ifs and buts, here's a good IF for you.
Masternodes are NOT the solution to decentralize everything effectively.
Here's the level of decentralised attack-(in)vulnerability that blinded Masternodes are capable of:
How can one say they are 100x more secure now, or 90x? I mean, where is this number coming from?
Probability of following Darksend through
- 4 non-blinded rounds with 10 masternodes* is (10/2300)^4 == 3.5734577849564574e-10
- 4 blinded rounds with 10 masternodes is ((10/2300.0)^20)**4 == 1.1528508353537067e-189
Each round uses 20 random masternodes of 2300, so you must control 20 of 2300 four times in a row. It's super secure .
Here's the new probablities for each successive round:
- 1 rounds with 10 masternodes is ((10/2300.0)^20)^1 == 5.826976675086318e-48
- 2 rounds with 10 masternodes is ((10/2300.0)^20)^2 == 3.3953657171999996e-95
- 3 rounds with 10 masternodes is ((10/2300.0)^20)^3 == 1.9784716837512123e-142
- 4 rounds with 1000 masternodes is ((1000/2300.0)^20)^4 == 1.1528508353537028e-29
* attacker controlled
Replace 'Probability of following Darksend through' with 'Probability of compromising the blockchain'... Masternodes can make even 'perfectly' decentralised solo mining obsolete as a security mechanism, never mind pooled mining. Evan has created the technology to nail the DRK network down on
all fronts, IMO we should use it. Why stick with something that's a thereisn'tevenawordforit-times more prone to attack?
I'm not claiming that exactly the same numbers would apply, but you get the idea - it's thousands of times more secure just from a number-of-servers perspective - add in blinded random subsetting and it's effectively impervious.
So yeah, imagine one person owning half the Masternodes, or three quarters, or 90%...
it wouldn't fucking matter.* Masternodes
are the best solution right now to decentralise everything effectively.
* eg. 90% compromised MNs = (((2070/2300)^20)^4) = 0.000218475
