Quote from: starpool on February 25, 2014, 04:56:36 PM

+1 for a donate button, or a way to contribute more. I think you're doing a great job. My hash rate is > 20 MH/s and I'd gladly donate .5-1% above your fee. Many of us came from MC and that fee was 3.88%.

Noted. I've had a few other people email me about this. My answer thusfar has been "I have more pressing stuff to attend to". I'd of course be happy about an added donation, but I don't feel right taking it until the pool is the best it can be. I've got things that I'd like to get done regardless of donation factors, but they're on my list eventually (easy answer is a "donate=0.5" password for a half percent donation - done some time in the future).

Quote from: CrashOD on February 25, 2014, 05:51:10 PM

I'm using minerd miners (no default stratum support) connected to a central stratum proxy. That stratum proxy connects to a headless linux VPS forwarding all traffic to whatever pool I choose. All I did was change the IP on the VPS to point to Middlecoin instead of Wafflepool and all of my rejects went away. So something else is going on, because my miner configurations haven't changed at all since I started mining LTC a year ago. I guess I was just having connection issues to just Wafflepool.

Absolutely not trying to call question to this, if MC is performing better for you, thats nothing I can argue with, and you should do due diligence to make sure! That said, minerd (cpuminer/pooler) has full support for stratum, and running it through a proxy will only negatively affect it. Please also make sure you're connected to the closest endpoint (closest meaning ping-wise, not necessarily geographical location).

Quote from: lagster on February 25, 2014, 08:56:51 PM

Quote from: poolwaffle on February 25, 2014, 05:06:33 PM

Quote from: lagster on February 25, 2014, 04:19:55 PM

Are you working on/going to implement soon some sort of "coin flipping" distribution for all the wafflepowah at your command over more than one coin at a given time?

We're actually testing it today on a more broad scale than we'd like (making it as fine-grained as we want is challenging on a technical side). Right now, about 1/5th of our hashrate (EU cluster) has a more open set of coins that it can deal with (essentially 5+ difficulty), where the others are only dealing with 20+ for now. We want to split it further than that, but again, its not something that happens overnight (unlike the hashrate boost).

Thank you for the answer and thanks for the work you are doing. I guess even different sets of coins for clusters will do the job and will let us get some profit from those low difficuly coins no matter how much more will wafflepowah grow.
And if this is the esiest way to spread gh/s around for the sake of using those <20 coins, would 1.5% or even 2% fee cover your expenses, to run 2-3 more clusters? You do a lot of math anyways, what are your estimates, will it lead to an increase in average profit for miners over today's 0.0098?

Quote from: richmke on February 25, 2014, 09:30:51 PM

Quote from: lagster on February 25, 2014, 08:56:51 PM

And if this is the esiest way to spread gh/s around for the sake of using those <20 coins, would 1.5% or even 2% fee cover your expenses, to run 2-3 more clusters?

The problem is throwing too much hashing power at the small coins. Adding more clusters will throw more hashing power and overwhelm it. A better option is being able to somehow fine tune how large the cluster is that is thrown at it.

Grouping these two together.
Unfortunately, its less a matter of getting more clusters. Although more clusters breaks us up into smaller distinguishable groups (instead of having 4 USEast servers, we might have 8-10), it doesn't get rid of the problems of the USEast servers been 200-300ms away from our EU cluster (regardless of how large that cluster is). Although I appreciate the (assumed) sentiment of a larger fee % would afford more servers would assume slightly larger earnings, I think a larger part of it is just playing with our algorithms and getting things tweaked for where we are currently. Paying 3% fees, for a 4% boost is obviously something everyone would agree to (yay math), but I think we can get a 5-10% boost without adding the fees yet, and I'd like to focus on that first

That said, if it comes down to it later, and I can reasonably-well promise and increase in earnings compared to an increase in fees (hiring someone, more servers, etc), I'll definitely bring it up when I have hard numbers.

The second part, throwing too much hashrate at small coins, is an issue I've mentioned in the past, and I think we have a good solution for it. That said, I don't want to mention the whole thing here (I think its at least minorly innovative), and ask for a few more days of time to code/test it out. The unfortunate thing here, is I can code up a new idea, wait 4-6 hours to see some results, and have the results skewed by luck pretty heavily still. So it takes a few full days to test a new branch of code fully. Sucks, but its the nature of the beast. And as long as we're in line with the competition, there will be some days where our testing underperforms (bad code, bad luck, etc), and days where we overperform (good code, good luck, etc), and its unfortunately, very hard to tell the two apart.

Quote from: Sinyk7 on February 25, 2014, 11:21:34 PM

I gotta admit, I never had problems with waffle. Really low reject rates too.

I also admit that I am testing out clevermining to see if their claims of profitability match their stats. My reject rate there though.... Cry

Please report back, as this is a great piece of info. We can obviously claim 10% higher stats, if 10% of everyones hashrate gets rejected

I'm personally too involved to give any sort of non-biased stats, and I appreciate anyone that can come back and say "from time A to B, I earned C on wafflepool, and D on ". WP has things to learn from other pools, they have things to learn from us