Topic: [ANN][The Original Multipool - Scrypt/SHA256/Scrypt-N/X11] multipool.us - page 206. (Read 424294 times)

Sorry.. Vicurex.

and we are mining FTC now ...?
Maybe should switch to a LTC maybe while we find out?


Cryptsy has temporarily suspended FTC trading while he investigates.. according to Admin there.

Not suspended, but if there is a successful 51% attack, there is a chance your transaction could get reversed.

So I'd say "withdraw at your own risk".

Vic?

ftc WDs suspended?

Vic is still accepting..
T.

Hmm, maybe I should move some of the pool's FTC reserves into LTC for now.  Of course, that's assuming I can even get them into an exchange at this point.

I'm just kind-of glad I traded almost all mine few days ago.. just never know with kids these days..

T.

Actually, I rescind my previous comment.  FTC is currently being attacked.

We will most likely move to it in ~5 hours when the difficulty adjusts down, but I will monitor the situation and disable FTC if the attackers succeed.

Ahh lol ok.. (paranoia sets in..) I was in the chatbox on cryptsy and Vern had disabled ftc trading and so blah blahh...



T.

To expand on this (because if read a certain way, it seems to contradict kslavik's previous statement that PPLNS is always fair), PPLNS can be unfair if the number of shares is set too high because if there is a difficulty adjustment in the middle of the round, shares mined before the adjustment become worth more.

Flound.. you need to scrap FTC.. see http://imgur.com/a/fTye4/noscript#0

Bad news! Apparently the blocks been 51% attacked a few times...

T.

glad to see we are getting some extended time on PXC

I guess you are right on this.
As you said, only in extreme cases will it become unfair but those extreme cases will most likely never happen on multipool since LTC gets a healthy amount of steady power.

People who stayed inside LTC for the last 6 hours do deserve to get most of the reward. Because users in MP were getting rewards for other coins during those 6 hours. Reward is still distributed fairly based of # of shares contributed in the last 6 hours by all users.

PPLNS is fair for any amount of time or any # of shares as long as it stays the same. You cannot say that it fair if is set at 24 hours and not fair if it is set at 5 minutes. Both ways are valid and both are fair.

PPLNS only becomes unfair when there is a difficulty adjustment in play. But it only unfair if PPLNS is set at high # of shares or long amount of time.

Due to the nature of this pool PPLNS should be set to the shortest amount of time so the slowest user on the network would submit statistically significant # of shares so when the block is found, the reward would be distributed fairly.

This is true if the hashrate drops to 0, but if a few dedicated miners stay on the port, they could take the whole bounty after 6 hours by submitting very little shares in that time. I think the multipool miners should get some of that since they put in a lot of shares not so many shares ago (time ago yes, shares ago no).

If the time for the time-based PPLNS is set to 24 hours, the steady LTC miners will get their share on blocks found, based on their last 24 hours of mining while the multiport miners get a relatively lower cut on blocks found while they are on LTC since they have not been on for 24 hours straight. So yes, then it is fair that they also get shares on LTC blocks found while they are on a different coin for the next 24 hours.
It will most likely even out in the end anyway, and no matter what system you choose, the differences will be minimal over time I guess. I just wanted to share my thoughts on this and will stay on your pool whatever system you use since you did a really great job and like you said, the last day we made 3 times more than we could on bitcoin

This is true but remember multiport miners hopped onto another lower-difficulty coin and are getting paid for that.  LTC miners are staying on LTC and will only get paid if the pool finds a block.  And if the dedicated hashrate on LTC disappears and everyone's round shares go to 0, then everyone is starting on an equal footing when the multiport comes back, so I think that is still fair.  Remember, as kslavik said, that the idea of any kind of PPLNS is not that you get paid for every share you submit, it's just that block rewards are distributed fairly when a block is found.

Should the multiport miners get paid for their shares on LTC for a full 24 hours and also their shares spent mining other easier currencies for which they got steady payouts?  I don't know that that's necessarily fair.

But really, I don't see this being an issue except maybe in a corner case where the pool is mining LTC, has bad luck, then hops to NVC and continues to have bad luck and then does not find a NVC  block for 4-5 hours.  But this is a very low probability outcome.

In a case where the pool hops to any other coin, you are generating guaranteed profits again almost instantly since the block times are so low on those coins.  Plus you still have the potential of a LTC payout if the dedicated LTC miners find a block.

Even with the mediocre luck and the short multiport stints on LTC in the past day I have a great profitability, over 3x what I would have made mining bitcoin.

Why time-based PPLNS is a BAD idea for multipool.

I know there already has been a long discussion in this thread about what payment system would be best for this type of pool and finally Flound has settled on time-based PPLNS.
I'm not trying to restart that discussion so we can hear the same arguments again and again but I have given it a lot of thought and I think I have some points that are really relevant and have not been made yet.

Flound argues that for the LTC pool 4-6 hours would be a good period to set for the time-based PPLNS. This can be fair based on his assumption that the LTC port has about 100MH on it when the multiport is on another coin.
But (for argument sake) let's  look at the situation where the steady hashing power on LTC would drop to a low value or even disappear. If the multiport then moves away from LTC and doesn't return within 6 hours, it will be a FACT that all shares will be lost and the multiport miners will get nothing in return for their time spent on LTC.
So a low steady hashrate on LTC will be bad for multiport miners when using time-based PPLNS while on the other hand the 6-hour time-based PPLNS system will be unfair to the steady LTC miners when there is steady hashrate much higher than 100MH.
Based on the above, you would have to adjust the time period of the time-based PPLNS system, based on the hashing power that is on the port. Which boils down to ...... (think about it) ...... exactly the same a regular 'share based' PPLNS.

Then we have a lot of people claiming PPLNS is not suited for multipool because it was specifically designed to prevent pool hopping. While the last part is true, the multiport is not pool-hopping in a regular way.
Miners that use pool-hopping to increase their profit, hop on a pool when that pool has just found a block and starts a fresh round with 0 shares taken so far. They stay on the pool for a while but if the round is not 'lucky' and takes too long, they hop onto another pool that has just found a block and started a fresh round. PPLNS is designed to prevent this by invalidating the oldest shares on unlucky rounds.
The multiport does NOT hop onto a coin because the pool has just found a block and starts a fresh round. The multiport hops on a coin because the difficulty and/or exchange rate has changed and it became the most profitable to mine in that regard. So the point where the multipool hops into a round compared to when that round started, is absolutely random. It can be early into the round but it can just as well be at the end of a round that has been very unlucky.
For this reason, share-based PPLNS will have NO negative effects for multiport users and the effects of lucky/unlucky rounds WILL even out over time.

Actually, share-based PPLNS is even better for this type of pool than a Proportional system. With a proportional system, it could be argued that the multiport would be really 'stupid' to jump onto a coin when there is a really unlucky round going on for a long time already and many proportional shares have already been mined (unless the coin would become very profitable by a really big margin ofcourse). Share based PPLNS also evens out this effect so as far as I can see it, this is the perfect system to use on this pool.

Maybe I'm missing something here and my reasoning is flawed (I'm only human after all) so please share your thoughts if you catch something that I didn't. I'm always eager to learn.
At the end of the day it will be Flound who decides what system to use but I hope he gives this a little thought

Very nice! Great work!!!

A small suggestion... in the top left table it would great to have the balance in bitcoin equivalent.

You should also implement a fixed fee to use the multipool, maybe 2-5% since it's what's taking most of your time. You could leave the direct mining free.

I noticed that Franko and Megacoin are often more profitable than the coins on the multipool so it would be great if you could add them. These coins have a much higher average network hash rate than MNC, LKY ARG or PXC so would see less dramatic difficulty swings when the pool would hit them (well, at least less than the previously mentioned coins).

I thought about doing that but it's a lot of data to save that for every user though.  I'll consider it.

I like it!

I have some other thoughts for consideration of making it onto your to-do list. I realise none of this is essential - will just add more to your workload - but I figured may as well jot down what I was thinking..

For the two tables on the top left and right - I think it would be nice to have clickable sorting options on the column headers - to change the order to your preference (e.g. alphabetically, or in order of profitability). 

Also, instead of having a pointer to the coin that the Multi Port it currently mining, what about just highlighting the relevant row? And if necessary, you could add a note / explanation below the tables, or in the space between them below the "Currently Mining" box. And you could then use that space again on the right-hand table to rather list the static mining port numbers for each.

And on the My Stats page for the newly added list of payouts, could we get the relevant timestamp for when each block was mined?

Then the new Profit Snapshot is really great. I think that could warrant some further historical graphing.. showing profitability history/ trend. That page could be a separate tab on its own, e.g. "My Profitability" next to "My Stats".. with a historic graph below the 24 hour snapshot table. You'd have to put a stake in the ground for the time of day at which each point is plotted on the graph, but it could be great to see. And that leads to thoughts about a comparison / ratio vs the pool as a whole, and/or between individual coins..

ok, ok. Enough already!