[4+ EH] Slush Pool (slushpool.com); Overt AsicBoost; World First Mining Pool

[slush]

It's absolutely fine now

Damn, so that magic "bug" is solved. Many users were complaining for pool performance and "magical network issues" (mainly GUIminers, now I understand why!). And it was because they were using bad URL and mining thru webserver. I setup this transparent proxy 6 months ago right because GUIminer's hardcoded old URL, but then I completely forgot to it. And I was unable to link people's mining issues with webserver load, LOL.

[worldinacoin]

I better give it a try with the new url again, the namecoins are too alluring.

[naypalm]

Good luck with the migration, hopefully it goes smooth and problem free!  

[AngelusWebDesign]

Hey Slush -- I have a question about your algorithm.

Say you have a 24 hour round (purely hypothetical!) and my miners are with you for the first 12 hours, then I switch to another pool.
My BTC reward will dwindle to almost nothing; that much I understand.

But what about the Namecoin blocks?  Say you find 50 the first 12 hours, and 50 the second 12 hours. Would I get my usual share of the first 50 (the time I was contributing fully), and then next to nothing (or actual nothing) of the "second" 50 NMC blocks?

In other words, are NMC blocks divided up based on the shares for that NMC block, or is some other algorithm used? I guess I'm asking if NMC and BTC division are separate.

I would hope so, otherwise anyone hearing that 50 NMC blocks were found on Slush would hurry up and join the round, and contribute full shares for 5 or 10 minutes, to get a huge payday!

Thanks,

Matthew

[AngelusWebDesign]

It's looking like my guess was incorrect -- it looks like you calculate ALL the NMC blocks proportionate to the last BTC round.

So if you get a long round (like the last one -- 12:58) and word gets out that there's a nice NMC jackpot, everyone's going to want to join for that last 30 minutes so they can get a massive NMC payday.

That's ridiculous if you ask me.

I, for example, was with your pool 100% for the first 7 HOURS or so of the round -- I should get my usual share of NMC blocks found during that period.

[Bigpiggy01]

The only problem with what you're saying above is that no one really knows when a block is going to end. Reverse hopping is only theoretically possible as no one really knows how long the block will be 

[cosurgi]

... everyone's going to want to join for that last 30 minutes so they can get a massive NMC payday.

lol.  

Yeah, everyone should join! (let's keep it a secret that it is unknown when are the last 30min).
People, people - huge jackpot here!

[bulanula]

Is the France migration finished now ? Thanks.

[Meni Rosenfeld]

The only problem with what you're saying above is that no one really knows when a block is going to end. Reverse hopping is only theoretically possible as no one really knows how long the block will be  

AngelusWebDesign is right - assuming the payment is done as he described, and even though he did not articulate it perfectly.

If the round is old and the accumulated NMC reward is high, it is more profitable to mine.

... everyone's going to want to join for that last 30 minutes so they can get a massive NMC payday.

Yeah, everyone should join! (let's keep it a secret that it is unknown when are the last 30min).
People, people - huge jackpot here!

No free lunch. If it's more profitable to mine in an old round, it means it is less profitable to mine in a young round.

[RobertRibbeck]

Is the France migration finished now ? Thanks.

Don't think so the web site & jason aps are down

[slush]

Don't think so the web site & jason aps are down

No, it's up. Refresh your DNS...

[DutchBrat]

The only problem with what you're saying above is that no one really knows when a block is going to end. Reverse hopping is only theoretically possible as no one really knows how long the block will be  

AngelusWebDesign is right - assuming the payment is done as he described, and even though he did not articulate it perfectly.

If the round is old and the accumulated NMC reward is high, it is more profitable to mine.

... everyone's going to want to join for that last 30 minutes so they can get a massive NMC payday.

Yeah, everyone should join! (let's keep it a secret that it is unknown when are the last 30min).
People, people - huge jackpot here!

No free lunch. If it's more profitable to mine in an old round, it means it is less profitable to mine in a young round.

Isn't that always the case with the score-system ?

The oldest shares use their value after a while, half an hour after you stop contributing to a round your already contributed shares have lost all their value, that was always the case. so the longer the round takes the more incentive people have to get 15 mins worth of mining in to be credited with the same reward as people that have been contributing since the start of the round. With the scoring-system it makes no sense to stop mining in the pool when the round takes longer and longer, since you will lose all the work you have done in that round....

As long as the auxiliary chain is not 'scored' as a seperate chain, (slush said he is going to change the pool in the future to make that happen) the NMC reward is based on your BTC-'scoring'

[organofcorti]

No free lunch. If it's more profitable to mine in an old round, it means it is less profitable to mine in a young round.

Isn't that always the case with the score-system ?

The oldest shares use their value after a while, half an hour after you stop contributing to a round your already contributed shares have lost all their value, that was always the case. so the longer the round takes the more incentive people have to get 15 mins worth of mining in to be credited with the same reward as people that have been contributing since the start of the round. With the scoring-system it makes no sense to stop mining in the pool when the round takes longer and longer, since you will lose all the work you have done in that round....

As long as the auxiliary chain is not 'scored' as a seperate chain, (slush said he is going to change the pool in the future to make that happen) the NMC reward is based on your BTC-'scoring'

I think you're mixing up two different things.

The expected value of a share in this pool is always highest in the early part of a round and then is maintains at just above 0.9 (atm), as you can see in the following graph:

(although  payout system based variance increases with increased round time that is not shown here)

If there is a larger reward than usual because a round is longer, then this changes the payout to favour shares in longer rounds since the value of each share increases.

[DutchBrat]

That is completely true,

I guess what I meant to say was:

If you have a long long round like yesterday and you would have stopped mining at hour 12, you would have gotten NO payout at all, whilst when you would have started at hour 12, you would have been paid exactly the same as people that had been mining the entire time (or am I wrong in assuming that? - I had a disconnect for 40 mins during that round because I was directing my miners to the website instead of api.xxx.xxx and my estimated reward dropped to zero, in about 30 mins after I had reconnected my estimated reward was back to normal).

So it makes no sense to drop out of a long round, since you will have mined all that time for nothing, but mining 12 hours somewhere else and then mining here would have enhanced your reward.

But that is only theoretical, since you can never know how long a round will last

And luck evens out in infinity anyway, so keep hashing away at Slush's pool ! 

[organofcorti]

So it makes no sense to drop out of a long round, since you will have mined all that time for nothing, but mining 12 hours somewhere else and then mining here would have enhanced your reward.

But that is only theoretical, since you can never know how long a round will last

And luck evens out in infinity anyway, so keep hashing away at Slush's pool ! 

Indeed!

However, a small correction - never knowing how long a round will last is not a theoretical problem, it's the basis of probability. If you did jump in at the 12th hour you'd have a good payout, but 300 seconds later your payout is diminishing exponentially. If you can't guess the end of the round precisely, you can't game the pool at the end of the round.

[Meni Rosenfeld]

So it makes no sense to drop out of a long round, since you will have mined all that time for nothing, but mining 12 hours somewhere else and then mining here would have enhanced your reward.

That's a common myth. If you've mined for 12 hours and no block was found, then in hindsight, you have mined all that time for nothing whether you drop out or continue mining. Your future mining has no impact whatsoever on what you get for the past mining done, you're faced with a new decision on whether the potential rewards for your future mining are worth it or not. In a hopping-proof method the profitability is always the same.

[cablepair]

Mine at a SMPPS pool and then you will love the longer rounds

[organofcorti]

Mine at a SMPPS pool and then you will love the longer rounds

Mine at an SMPPS pool with longer rounds and it will take a long time to get your coinage out. I've been waiting for 6 weeks for some coins from bitpit, an ESMPPS pool.

Mine at an SMPPS pool that's had a bunch of shorter rounds and your time to payout is reduced.

[naypalm]

I remember mining at bitpit a long time ago, I was getting 3 or 4 btc payout, but it took weeks to finish the round and forever to get paid.

[slush]

Reward from current round #8628 will be calculated wrongly. I'll fix it manually later.

Current status:
* Website moved
* api.bitcoin.cz moved
* api2.bitcoin.cz still in London
* database still in London

If you see some performance problems on api.bitcoin.cz, tell me.

To be continued...