I hate to be a Negative Nancy...
- Thread starter camosoul
- Start date
camosoul
Well-known member
Yeah... But I've used that handy-dandy ban tool they built into it to ban everything that's not 12.1. I'm maintaining connections only to 12.1 nodes.the network is still updating - mixing will not work until that is stable and x amount of MNs are on 12.1 to take care of this business for you
it is 3 days after release - u know the drill
But, yeah, I guess I am a bit early to the party... We'll see how things go with more consensus in a week or so.
I did notice there's 12.1 with two different protocol versions...
flare
Well-known member
In contrast to
Let network transition to 12.1 first ...
I've just kicked off a 2 round 100 Dash privateSend mixing session... Let's see how long it takes...
Update: 15mins in and we're 50% done. Impressive
Edit: finished within 30mins. That's a big improvement on 12.0. Really impressive!
Walter
Let network transition to 12.1 first ...
Seems like there are plenty of masternodes on 12.1, so I figured I'd try some mixing. I went with 220 for 3 rounds. Shot up to 180 finished within 2 hours, the rest is slowing down a bit. With our fancy new "keys left" warning light, I can say that I have burned through 1030 keys so far on this mix.
------------------------------
EDIT
After further review, the 10s and the 1s are finished. The holdups are the .1 and especially the .01s. This might take a while as there probably aren't a lot of .01s out in the mixing bowl.
-------------------------------
Several hours later...
lowered rounds to 2
209.5 Dash in Privatesend
1400 keys used.
I stopped mixing.
------------------------------
EDIT
After further review, the 10s and the 1s are finished. The holdups are the .1 and especially the .01s. This might take a while as there probably aren't a lot of .01s out in the mixing bowl.
-------------------------------
Several hours later...
lowered rounds to 2
209.5 Dash in Privatesend
1400 keys used.
I stopped mixing.
Last edited:
I'm surprised so many people are mixing already. That sounds good, like people are availing themselves of this feature more often now
camosoul
Well-known member
It's running faster than it was since the MN count ticked over 55%. It tends to speed up and slow down considerably. I get bursts of tons of mixing, and then long periods of 12.0 speed. I'm even running as my own 3-way liquidity provider just to check and see, and that pattern persists. Interesting, but, still, we're way shy of the version consensus we need to call the results meaningful. 0.12.1.1 seems to be dipping it's toe in the pool according to version reporting...
I've been mixing since release with even numbers on 3 separate clients, and I'm just barely 20%; with most of that in the last 2 days. It's a hell of a lot better than it was, but it's still not "fast."
I've been mixing since release with even numbers on 3 separate clients, and I'm just barely 20%; with most of that in the last 2 days. It's a hell of a lot better than it was, but it's still not "fast."
camosoul
Well-known member
I've had the opposite experience... PS seems to have a balanced view of the quantity of denomination with disregard to value.
If you've got 100 DASH to denominate, and then mix it 8 times... You get pretty far pretty fast on the 10s. But it takes an order of magnitude longer for each decimal place, because they are given equally weighted presence in the submission engine.
Now, if you've got 100,000 DASH, even the 10s take forever, and the .01s, ugh, forget it... With 4 denominations, that means 25,000 has to mix 8 times doing it only .01 dash at a time...
You can split it into 50,000 on two clients without breaking anon, but it's still a bad idea, and you're really just providing a chance for more matches, it's still breaking down at 12,500 each in .01s... The mix acceleration impact is minimal and definitely not worth what you're giving up by being a single entity in 2 of the 3 parties present. But, then, again, who would expect you to do that with that information in hand...
I'm mostly just doing this to satisfy my curiosity. I doubt these analytics have any value to anyone but me.
Last edited:
camosoul
Well-known member
This does bring about an interesting idea tho... What if the 3-party mix was a minimum, and if the opportunoity to do a 3+N party mix presented itself in an equally timely fashion as 3, the option/opportunity to mix with 3+N parties would be taken. This would add more entropy. Simply observe circumstances and take action when available. Passively opportunistic. It probably wouldn't happen much, but it'd monkey-wrench the presumption that it was always 3 from a data farming perspective.
Now that it's faster, more people might use it, thus, presenting a feasible argument that such a feature could be useful.
Since mixing is still somewhat based on blockrate, you could deliberately augment it... It just piles more into each block with multisession... Since blocks virtually never happen within 20 seconds of each other, you could use that as a deliberate delay... PS spends 20 seconds after each block on pause, looking for matches for a 3+N. It holds and matches 3+N. Then continues with the FIFO way of doing it that it currently does. That'd be pretty cool...
Now that it's faster, more people might use it, thus, presenting a feasible argument that such a feature could be useful.
Since mixing is still somewhat based on blockrate, you could deliberately augment it... It just piles more into each block with multisession... Since blocks virtually never happen within 20 seconds of each other, you could use that as a deliberate delay... PS spends 20 seconds after each block on pause, looking for matches for a 3+N. It holds and matches 3+N. Then continues with the FIFO way of doing it that it currently does. That'd be pretty cool...
Last edited: