Just going to point out that this is how most Bitcoin ponzis and scams have started.  Somebody with "another business purpose" which they won't discuss in any detail, and out of the goodness of their hearts is deciding to share some of their profits with everybody else for no reason at all.

Not saying that is what is happening here, but you should *NEVER* join something where there is a too-good-to-be-true reward (110% PPS = the operator is losing 10% *OR MORE* in the long run for every share you submit) with the explanation for its existence being "another business purpose".

It should work just fine?  I don't offer any support for specific devices, but there shouldn't be anything special needed.

Now that the forums are back:  There were 2 sets of rolling restarts earlier today, with each pool server going down for about 5 seconds, though there were always 6-7 pool servers still online to failover to during the restarts.  Most users probably didn't notice the restarts, and the pool's stats before and after restarts seems to confirm that most miners these days will no longer stop functioning properly/failover permanently like they had a tendency to do in the past.

These restarts updated 3 things on each mining server:
1) Updated bitcoind [BTC Guild was a few updates behind, prefering caution to fast upgrades after what happened with 0.8].
2) Updated bitcoind settings to 900,000 byte max blocksize.  BTC Guild previously had this value set at 500KB.
3) Updated to a faster version of Matt Corallo's Relay Network on all servers (some were still using a python version).

731.6 * 1024 ~= 750,000 bytes [slightly less, but it's hard to make a block exactly 750,000 bytes].  It's a set of pools which set their maximum block size to 750000.

A couple months ago the botnet that was being used to try to brute force logins using a database of leaked username+password combinations from other sites got smarter and changed their tactic from trying to log in to trying to register new accounts with usernames from that database, basically letting them know if they at least have half of it right.  As a result, the User ID's got bumped up by about 900k overnight (a few months ago).

Still not at 1 in 1,000 levels of bad luck yet.  Given slush's stats imply the pool has had ~23,000 rounds (stats say 23414, but I think there's been a few gaps due to db sync issues/bugs/tweaks over the years), it's unlikely this is a record breaker in terms of CDF (hasn't slush had a 99.99, AKA 1 in 10,000, once in the past?).

EDIT:  Now at 1 in 1,000 (or worse) levels.

Going to be an interesting time going forward.  I'm curious to see whether this huge price decline is going to further centralization to private farms (properly located huge mining farms can keep running at a profit WAY beyond *most* individuals), or bring the balance back towards individuals and pools which are made up individual miners.  So far, it looks like BTC Guild is growing a bit [hashrate and network % both] even though it's quite clear that a number of farms and miners are turning off.  But that may just be people hopping onto the pool after the lucky few days.

Current CDF according to slush's stats is 99.6%.  That means you'd expect a block this long 1 out of every 250 rounds on average.

EDIT:  To clarify, this is a counter to the "something is fishy/something is up" idiots above.  While 1 in 250 might sound rare, that's probably in the once every 4-6 weeks range [on average].

Or any exchange for that matter.  I wonder how many people were (or are still) mining on Eligius using a Bitstamp deposit address that is no longer recognized.

EDIT:  Actually, I don't think Bitstamp worked with generation txes anyways, so probably 0.  Still, don't use any exchange as a wallet.

A few (3) payments failed to process due to the spike in luck yesterday.  Since BTC Guild credits your account after 1 confirmation, but the pool can't actually send funds from those blocks until they have over 100 confirmations, the big spike in luck caused the pool's hot wallet to be emptied of spendable coins until more confirmations came in for our recent blocks.

The failed payments should re-issue in the next 30 minutes.  This *may* happen again due to the luck spike we're having today, it all depends on how many accounts cross an automatic payment threshold at the same time vs the timing of our blocks becoming spendable.

The pool always has full funds to pay all users balances, but only a fraction of those funds are kept in the hot wallet, while the rest are moved into cold storage.

Unless your hardware is having problems, leaving it at default is fine on every(?) pool.  I'm fairly certain every pool offers variable difficulty these days, so your hardware is automatically adjusted to the optimal difficulty.  The only time this doesn't work is with certain hardware which is unable to function at lower difficulties properly, or ignores the pool provided difficulty.  You would know that's happening if your hardware is SIGNIFICANTLY under-performing according to the speed estimate on the pool interface.

It's best practice to split each physical machine to its own worker.  This allows you to check each worker's performance on the pool interface and your local interface.  It makes pool-side graphs significantly more useful for analyzing potential problems (miner slowing down during the hottest point in the day consistently -> problem with cooling).

It also means, on pools that offer it, you can receive notifications from the pool telling you that a worker has stopped mining.  If you combine multiple machines into one worker, the pool has no idea unless all the machines stopped simultaneously.

That's gambler's fallacy/observational bias.  Had you left after the 200% luck spike, you would have missed out on consistently > 100% rewards for the following 12 hours, and what will be *another* 200% [or more] luck spike once the next 3 shifts close.


EDIT:  The chart is going to have a record high spike in luck once the next 3 shifts close.  As of this post there are 3 shifts with 12 blocks paid out to them already, and all 3 are still open to being paid more.  The previous luck spike was a record and that was from a set of 3x 11-block shifts in a row.

*AND* broken/inaccurate even for a year ago.  Eligius/Guild never had a dive like that, and Slush hasn't been bigger than BTC Guild since 2011 yet the chart shows Slush on top at the end of it with Guild & Eligius not even visible.

Because the graph is inaccurate, simple as that.  There was never a nosedive in hashrate of BTC Guild or Eligius, so the graph is clearly being built off of inaccurate data.

That's why I said assuming luck always balances out to neutral.  Probably should've included quotes around "max possible".

94.5% would put you at equivalent of a 4.5% fee pool (assuming 1% orphan rate, unpaid orphans, and neutral luck) or a 5.5% PPS pool.  But since you can't know what the luck will be in advance, there's no way to know until after it has already happened.

Not sure where you got "97% is about the max possible".  The rule of thumb orphan rate is 1%, which would put the max possible (assuming luck always balances out to neutral) at 99%.

Targeting a specific address wouldn't matter.  The odds are so astronomically high that I can't even think of a statistical analogy.  Probably something along the lines of winning the powerball lottery hundreds of times, in a row, with only a single ticket for each lottery.

Yes, it can happen.  But at some point you have to just accept that it won't ever happen in the lifetime of the universe, unless a flaw is discovered that allows somebody to reverse the process of creating an asymmetrical key pair.

Satoshi made mention of the fact that the process of mining Bitcoins would become centralized over time as the resources needed would exceed what the average user could dedicate.  It's doubtful that the idea of pools would "scare Satoshi away" when he knew that mining centralization could occur.  Pools were a preferable alternative compared to consolidation of power among just a few large companies that controlled the hashing power directly.


[I'll try to find the quote directly which dealt with mining centralization]


EDIT:

I'd bet that the majority of miners are in China these days, it makes sense that Chinese pools ended up exploding in popularity.  Back when BTC Guild was hitting 40%+ of the network when ASICs first hit the scenes, the site was getting more than 50% of its web traffic from China alone.