Author

Topic: KanoPool kano.is lowest 0.9% fee 🐈 since 2014 - Worldwide - 2432 blocks - page 1191. (Read 5352229 times)

legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
moving kano's pool payments...
1.5 BTC equals about 17000 bytes for me a.k.a lots of bytes.
I have been playing to see what is the minimal sat/byte network will accept and confirm within, say, 4-6 hr. 10 min confirm is unneeded.
Of course, 21co rec 100sat/byte, but I recon that even 15-20 sat/byte for a large byte size transaction would do (at least a couple of days ago).
It works for me and saves basically 80-85%.
What is your experience?
Well if you haven't actually sent it yet, you could try lower and see if we pick it up in the next block.
As long as it gets to (and stays at) the pool, the pool should confirm it, but I guess the problem may be if other bitcoin nodes will at least relay it.
Zero fee usually wont get relayed, but I'm not all too sure on the general 'average' rules bitcoin nodes use to relay transactions to other nodes.
sr. member
Activity: 276
Merit: 250
Yay! Good start to the new day.  Grin
BTW, hate to admit it but I had to look up definition of "Organ of Corti". Guess I forgot most my high school biology. Man, the inner ear is one heck of an amazing piece of biological technology! I'm not smart enough to have a handle called "OrganOfCorti". If I had to pick a handle based on part of our hearing I probably would just be "Ear Lobe". Hahaha. Mine on!

[EDIT] @Organofcorti: Just to be clear, no way do I mean any disrespect to you. On the contrary, I am complimenting you as I assume you are a super smart person, just like Kano. Intellect wise, no way I can keep up with you guys.
legendary
Activity: 952
Merit: 1003
legendary
Activity: 3892
Merit: 4331
moving kano's pool payments...
1.5 BTC equals about 17000 bytes for me a.k.a lots of bytes.
I have been playing to see what is the minimal sat/byte network will accept and confirm within, say, 4-6 hr. 10 min confirm is unneeded.
Of course, 21co rec 100sat/byte, but I recon that even 15-20 sat/byte for a large byte size transaction would do (at least a couple of days ago).
It works for me and saves basically 80-85%.
What is your experience?
legendary
Activity: 952
Merit: 1003
I moved my S9's as well. Although there isn't any hard proof that S9's don't do as well on Kano, I didn't want to drag the pool down. I'd much rather bring my 300th back

I think that's been my point, Sierra...and I also think you folks with the blocks of S9s are doing a good thing, even if the pool rate suffers for a bit...but it does seem, at least in this short(er) term, that we're not doing so badly. Y'all will be back, I tink.  Cool
sr. member
Activity: 546
Merit: 253
I moved my S9's as well. Although there isn't any hard proof that S9's don't do as well on Kano, I didn't want to drag the pool down. I'd much rather bring my 300th back
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
Well since it's got to the extreme level of even organofcorti commenting on it, I'll post the magic numbers for citronick Smiley

All history:
23.51741743 BDR : 26 Blocks - good! Smiley

Since the halving (i.e. a subset of the above numbers):
11.32364258 BDR : 11 Blocks - no problem at all Smiley

i.e. he's found more blocks than expected, and since the halving, very close to expected.

I've not really said anything about it since my post on the Bitmain S9 stats, since as I said there:

Kano it would be interesting to know if s9's are in some way having a negative effect on your pool. If they are it would be in the best interest of the pool that they go somewhere else including mine. Just curious.
Well for miner types directly connected to the pool (not via a proxy)

Since they first existed:
67.76844935 BDR for 60 blocks - so not excellent, but not impossible: CDF[Erl] 0.842606

Since the halving (excluding the data before the halving):
63.99816529 BDR for 57 blocks CDF[Erl] 0.825166 (thus a little better)
But since the halving there are worse - the longer term figures are (usually) statistically more relevant

(there seems to be a v2 bmminer that's found no blocks so far, but the BDR is only 0.07520162)

They do have the worst luck and the worst CDF on the pool for miner types connected directly to the pool.
But the numbers aren't in the realms of improbability - and it doesn't include any using proxies.

However there is (only) one "proxy type" with a worse CDF
stratehm-stratum-proxy-0.8.0 7.76724928 BDR 3 blocks: CDF[Erl] 0.983517

As I said above, that's only S9's connected directly to the pool, if they use a proxy then they wont show up in those numbers.
...
Quote
They do have the worst luck and the worst CDF on the pool for miner types connected directly to the pool.
But the numbers aren't in the realms of improbability - and it doesn't include any using proxies.
i.e. the number in that quote
Quote
... so not excellent, but not impossible: CDF[Erl] 0.842606

Edit: actually the updated numbers since that post - since there's been a few more bmminer blocks and of course a higher BDR
bmminer 71.24030662 BDR : 65 Blocks CDF[Erl]: 0.785699
bmminer2 0.19718556 BDR no blocks yet
hero member
Activity: 777
Merit: 1003
I think maybe everyone is thinking about this way too hard.

I may be wrong, but I think Kano already has a way of determining if a user is doing what they should within the statistically expected range. This is akin to the discussion awhile ago about block withholding.

If citronick's account shows that things are within the statistically expected range, he should be all set.
legendary
Activity: 952
Merit: 1003
Agreed...on both. I'm not knowledgeable enough on the particular subject to opine further. IOW, since I don't have a clue past seeing correlation, I probably need to shut up and just see what happens. I think eventually we'll know, even if we don't break our proverbial necks trying.

...and, stats was ALWAYS my least favorite area.  Kiss

EDIT: Spot's at $677...sumthin's up.
legendary
Activity: 4326
Merit: 8950
'The right to privacy matters'
I think everyone agrees with what Newko said.
citronick you really should come back, Juvia belongs here.  Smiley

Ditto!  Come on back citronick and let Juvia help you enjoy the next block party with today's spot price of $669!  Grin

Well...I'll third and fourth those tomes. Not to extend the S9 discussion...but, one of my students (a really brilliant person) in my grad school Research Methods class once quipped, "Yeah, correlation doesn't necessarily prove causation...but when you're observing a cat catch and eat a mouse, do you need to see ten other cats eat ten other mice to feel comfy in the statement that cats eat mice?" IMHO you did a good thing, citronick, and for the right reasons, whether it turns out that our paranoia is founded or not. For me and my teeny-tiny group here, we're doing OK still. No worries.
 



This is testable though, right? You run your mining software, record *all* the hashes 1. smaller than some size, and 2. for some amount of time that provides enough data 3. to test whether or not the data could belong to the theoretical population distribution. The test could be a "goodness of fit test" or use statistical inference.

"Enough data" might just mean enough hashes to determine whether or not citronick has the 32 bit error that some previous miners have had (though I assume that Kano has already run that particular test ).


yeah  I would really like for him to do some 5 day off 5 day on tests.

 he should hit a block every nine days.  so if he did say 3 tests of 9 days on and 3 tests of 9 days off.  it is 54 days I would love to see if for the 27 days he was on the pool was a lot different the the 27 days off the pool
donator
Activity: 2058
Merit: 1007
Poor impulse control.
I think everyone agrees with what Newko said.
citronick you really should come back, Juvia belongs here.  Smiley

Ditto!  Come on back citronick and let Juvia help you enjoy the next block party with today's spot price of $669!  Grin

Well...I'll third and fourth those tomes. Not to extend the S9 discussion...but, one of my students (a really brilliant person) in my grad school Research Methods class once quipped, "Yeah, correlation doesn't necessarily prove causation...but when you're observing a cat catch and eat a mouse, do you need to see ten other cats eat ten other mice to feel comfy in the statement that cats eat mice?" IMHO you did a good thing, citronick, and for the right reasons, whether it turns out that our paranoia is founded or not. For me and my teeny-tiny group here, we're doing OK still. No worries.
 



This is testable though, right? You run your mining software, record *all* the hashes 1. smaller than some size, and 2. for some amount of time that provides enough data 3. to test whether or not the data could belong to the theoretical population distribution. The test could be a "goodness of fit test" or use statistical inference.

"Enough data" might just mean enough hashes to determine whether or not citronick has the 32 bit error that some previous miners have had (though I assume that Kano has already run that particular test ).
legendary
Activity: 952
Merit: 1003
I think everyone agrees with what Newko said.
citronick you really should come back, Juvia belongs here.  Smiley

Ditto!  Come on back citronick and let Juvia help you enjoy the next block party with today's spot price of $669!  Grin

Well...I'll third and fourth those tomes. Not to extend the S9 discussion...but, one of my students (a really brilliant person) in my grad school Research Methods class once quipped, "Yeah, correlation doesn't necessarily prove causation...but when you're observing a cat catch and eat a mouse, do you need to see ten other cats eat ten other mice to feel comfy in the statement that cats eat mice?" IMHO you did a good thing, citronick, and for the right reasons, whether it turns out that our paranoia is founded or not. For me and my teeny-tiny group here, we're doing OK still. No worries.
 
Saw spot this morning as well...nice, steady, not-too-fast climb. Nice.

FWIW...I got a response from Canaan this morning. They've got a new ticket system, seems to work OK. They asked me to pop the box and take some photos; I'll do that later today...got some other things to take care of first. Hope to get something back online soon. Not even a whisper about an A7, of course. Sad

EDIT: Blatant commercial support here...Canaan's been really good overall about comms and following through. Hope they're able to "break out" soon.
legendary
Activity: 1736
Merit: 1032
Carl, aka Sonny :)
I think everyone agrees with what Newko said.
citronick you really should come back, Juvia belongs here.  Smiley

Ditto!  Come on back citronick and let Juvia help you enjoy the next block party with today's spot price of $669!  Grin
hero member
Activity: 777
Merit: 1003
I think everyone agrees with what Newko said.
citronick you really should come back, Juvia belongs here.  Smiley
sr. member
Activity: 276
Merit: 250
Well yesterday was a stellar day and woke up this morning to one in the basket. Grin. Very nice indeed!

@Citronic: Kudos to you for wanting to protect this pool from the S9. Very admirable. However, I do not think you have sufficient data to draw the conclusion that your S9s were causing the bad luck. Even if there were something fishy with your hardware, I agree with the others here who have said that at 1% of the total pool hash you're not going to skew the luck. If you want to keep your concience clear and if it's not to much of a PIA, try that 5 day on, 5 day off test for several cycles as Phil suggested and then look at the data (although you'd probably have to do that for at least 100 days to get a large enough sample size).

C'mon back, friend. You belong here and we miss lovely Juvia Smiley
Mine on!
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
member
Activity: 136
Merit: 11
Thanks for connecting the lines dots for me kano. That helps Smiley
legendary
Activity: 952
Merit: 1003
Correct. Just curious.
legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
Agreed, I am not sure I follow this (recent) conversation either....

firetreeactual was (I think) referring to who was making the 'unknown' blocks very recently using the ckpool pool software
(they've found 4 so far I think)

However the discussion before around page 1000 was not linuxjournal, that was gbminers
member
Activity: 136
Merit: 11
Agreed, I am not sure I follow this (recent) conversation either....

Jump to: