[OLD] Eligius: ASIC, no registration, no fee CPPSRB BTC + 105% PPS NMC, 877 #

[wizkid057]

i am thinking of let my both jupiter mine here. if i understand the CPPSRB system right it seems to be a nice payout method. the only thing i am afraid of is if i will earn the same amount of btc if i use elligius instead of btcguild (pplns) or 50btc (pps). 

Actually, long term you should earn more with Eligius than most other pools, regardless of their reward system, since Eligius has no fee.

-wk

[jgarzik]

i am thinking of let my both jupiter mine here. if i understand the CPPSRB system right it seems to be a nice payout method. the only thing i am afraid of is if i will earn the same amount of btc if i use elligius instead of btcguild (pplns) or 50btc (pps).  

Actually, long term you should earn more with Eligius than most other pools, regardless of their reward system, since Eligius has no fee.

Respectfully, that is not true as long as some pools give their users network transaction fees, and Eligius does not.

Transaction fees can add 1% or more to a miner's income.

It is important to pay some percentage of transaction fees to miners.  This provides virtuous economic signalling to miners and users alike.

I would propose moving Eligius to a rule like

• Takes X percentage of all transaction fees (10%? 50%?)  Right now it is 100%.
• Take no more than Y BTC in a single block, paying 100% of fees above Y BTC to miners.

[wizkid057]

i am thinking of let my both jupiter mine here. if i understand the CPPSRB system right it seems to be a nice payout method. the only thing i am afraid of is if i will earn the same amount of btc if i use elligius instead of btcguild (pplns) or 50btc (pps).  

Actually, long term you should earn more with Eligius than most other pools, regardless of their reward system, since Eligius has no fee.

Respectfully, that is not true as long as some pools give their users network transaction fees, and Eligius does not.

Transaction fees can add 1% or more to a miner's income.

It is important to pay some percentage of transaction fees to miners.  This provides virtuous economic signalling to miners and users alike.

I would propose moving Eligius to a rule like

• Takes X percentage of all transaction fees (10%? 50%?)  Right now it is 100%.
• Take no more than Y BTC in a single block, paying 100% of fees above Y BTC to miners.

As mentioned previously, there are several variations/experiments in the works on how to pay transaction fees to miners.  I had just gotten unexpectedly swamped with work to be able to finish the code.

With CPPSRB, paying transaction fees to miners is not completely straightforward or trivial.  There are many different ways to go about it, each with their own pros and cons and varying degrees of fairness.

In general, though, I believe I will be doing a variant of your list option #2.  Up to X BTC in fees to miners, remainder held by the pool.  This will let Eligius keep with its informal policy of returning accidental transaction fees to their owner (with signmessage proof for all inputs), minus X BTC.  After a certain amount of time has passed with no claim then the balance could be released to miners.  It is very obvious that any transaction with a large transaction fee is likely a mistake at this point.

I don't have my list in front of me, but, I will be trying several methods of fees payout (these are off the top of my head):

• 1. Add transaction fees to the 25 BTC rewarded to miners at the top of the share log when a block is found, thus paying more shelved shares using transaction fees.
• 2. Accumulate transaction fees in a fee-buffer of sorts.  Once X BTC is reached, buffer goes live.  The buffer would be paid to miners as a pseudo block when a miner submits a share that beats a target difficulty greater than 1/4th of the current network difficulty, but less than the network difficulty.  This way no one can know when the fees will be distributed.
• 3. Equally distribute block transaction fees to the shares in the share log that would be paid by the block that were submitted since the last block (essentially proportional distribution of fees to the current miners).
• 4. Add transaction fees to the block reward as calculated by CPPSRB and increase the PPS value for all shares paid by that would normally be paid by that block, new or old.

I've ordered these with what I feel would be the most fair first.  However, I will likely leave it up to the miners as to which system gets utilized long term, if any of these.

I do believe that long term Eligius will need a % of the transaction fees (NOT BLOCK REWARD) for expenses to survive.  For these experiments I will be letting the reward system distribute the full transaction fees accordingly, with the exception of the cap mentioned above.  I think long term a % of transaction fees would be kept each month until expenses are covered, then 100% to miners.

Personally, I feel that option one above is the way Eligius should go for fee distribution.  If at some point in the future all shelved shares are paid, then this could be reevaluated/amended accordingly, or fall back to option 2 or 3.  This would seem to be the most fair for Eligius miners, getting everyone closer to 100% PPS.

Edit: Also, just to point out, even if transactions average an extra 1% or so, as you say, pools with a fee > that % would pay out less than Eligius long term... and most larger pools have a fee > 1%.  Even a pool with 1% fee + 100% txn fees should be the same long term payout as Eligius.  But then you have to consider other factors, such as merged mining, where we pay 105% PPS, soon to be increased probably.

-wk

[The00Dustin]

Wouldn't option 3 encourage pool hopping?

[wizkid057]

Wouldn't option 3 encourage pool hopping?

Maybe, but realistically probably not.  Pool hoppers try to get in on short rounds for more reward on proportional systems.  With transaction fees, its likely that there would be less fees available to mine at the beginning of a round (since the pool should have just emptied its memory pool of transactions by mining the last block).  Also, transaction fees are such a small % overall, there would be little gain, if any, from hopping in the first place.

Also, I don't believe this is a best method for distributing fees anyway and listed it for completeness.

-wk

[jgarzik]

In general, though, I believe I will be doing a variant of your list option #2.  Up to X BTC in fees to miners, remainder held by the pool.  This will let Eligius keep with its informal policy of returning accidental transaction fees to their owner (with signmessage proof for all inputs), minus X BTC.  After a certain amount of time has passed with no claim then the balance could be released to miners.  It is very obvious that any transaction with a large transaction fee is likely a mistake at this point.

I do believe that long term Eligius will need a % of the transaction fees (NOT BLOCK REWARD) for expenses to survive.

Yes.  Set this up initially, when transaction fees are small, e.g.

• Eligius claims 10% of transaction fees, for pool expenses -- maybe even higher initially, like 50%, but plan to decrease percentage over time
• Plus a safety valve, if fees exceed 25 BTC in a single block (or pick your number)

1. Add transaction fees to the 25 BTC rewarded to miners at the top of the share log when a block is found, thus paying more shelved shares using transaction fees.

Yes, this is a nice option.

[mb300sd]

• Eligius claims 10% of transaction fees, for pool expenses -- maybe even higher initially, like 50%, but plan to decrease percentage over time
• Plus a safety valve, if fees exceed 25 BTC in a single block (or pick your number)

1. Add transaction fees to the 25 BTC rewarded to miners at the top of the share log when a block is found, thus paying more shelved shares using transaction fees.

Yes, this is a nice option.

+1

[driksson]

wtf

hashrate doubled over the last few days

correct, avalon has shipped a few batches, and they have been assembled. You see on the toplist of hashers ,a few accounts with 3TH each, which equals to one batch of 10K avalon chips.. "fun times ahead"..

oh ok i thought it was due to the 400gh units from bitfury and knc

Bitfury should be part of increase also. KNC only said they started shipping have ppl received units yet?

[amer]

wtf

hashrate doubled over the last few days

correct, avalon has shipped a few batches, and they have been assembled. You see on the toplist of hashers ,a few accounts with 3TH each, which equals to one batch of 10K avalon chips.. "fun times ahead"..

oh ok i thought it was due to the 400gh units from bitfury and knc

Bitfury should be part of increase also. KNC only said they started shipping have ppl received units yet?

I think it's refugees from the DDOS on BTCGuild and 50BTC. Those boxes haven't been widely distributed yet.

[wizkid057]

In general, though, I believe I will be doing a variant of your list option #2.  Up to X BTC in fees to miners, remainder held by the pool.  This will let Eligius keep with its informal policy of returning accidental transaction fees to their owner (with signmessage proof for all inputs), minus X BTC.  After a certain amount of time has passed with no claim then the balance could be released to miners.  It is very obvious that any transaction with a large transaction fee is likely a mistake at this point.

I do believe that long term Eligius will need a % of the transaction fees (NOT BLOCK REWARD) for expenses to survive.

Yes.  Set this up initially, when transaction fees are small, e.g.

• Eligius claims 10% of transaction fees, for pool expenses -- maybe even higher initially, like 50%, but plan to decrease percentage over time
• Plus a safety valve, if fees exceed 25 BTC in a single block (or pick your number)

1. Add transaction fees to the 25 BTC rewarded to miners at the top of the share log when a block is found, thus paying more shelved shares using transaction fees.

Yes, this is a nice option.

Fees to miners experiment number one has been implemented.

Currently 100% of fees are rewarded at the time the block is mined.  Unfortunately the fees portion can not be included in balance estimates accurately on the stats side, but are reflected when the block is rewarded.

This is implemented as of and including pool block height 261279.

While from this point on I will ensure that fees will be rewarded to miners in one form or another, this particular method is experimental and subject to change and tweaks as needed.  Eventually I do believe a % of transaction fees will need to be kept for pool operating expenses, but for now that is not needed and 100% goes to miners.

I will run this experiment for some time (few weeks probably) and evaluate its performance and reception with miners at that time and proceed one of two ways: Leave this method in place permanently, or continue on to the next fee distribution experiment.  All would be dependent on how well miners like the setup and how well it works in practice.

As of now, this should average roughly a 1% bonus to payouts using a fee average for the last 1000 network blocks.

Thanks for using Eligius! 

-wk

[2GOOD]

As of now, this should average roughly a 1% bonus to payouts using a fee average for the last 1000 network blocks.

Thanks for using Eligius! 

-wk

Great news. Thank You 

[wizkid057]

Continuing with pool improvements...

Soon, Eligius will no longer be serving any pool-difficulty 1 work, opting to a bump to pool-difficulty 2.

With the exponential rise in the average hash rate overall, difficulty 1 work is less and less useful.

I will be looking to make this change in the next few days.  This should have little to no real impact on miners, but will reduce pool-side load a bit.  It will also make it easier for higher hash rate miners to start mining.

Thanks!

-wk

[wizkid057]

Just to verify, stats page is down correct?

Incorrect. Stats page appears to be working fine.

[Rampion]

I've been mining at Eligius for the last 3/4 days and I love the interface and the 0% fees, but isn't the variance extremely high?

[TiborB]

I've been mining at Eligius for the last 3/4 days and I love the interface and the 0% fees, but isn't the variance extremely high?

I believe variance is high for slower minder due to the fact that min pdiff was increased from 1 to 2. At least that is how I explain my observation since it was bumped.

   T

[mb300sd]

Something broken on stats?

[Rampion]

I've been mining at Eligius for the last 3/4 days and I love the interface and the 0% fees, but isn't the variance extremely high?

I believe variance is high for slower minder due to the fact that min pdiff was increased from 1 to 2. At least that is how I explain my observation since it was bumped.

   T

I have a few TH/s and it looks to me variance is high

BTW, how do you cange min diff. to +256? It seems is not working on "my eligius" as I see cgminer mining at 127...

[Rampion]

I've been mining at Eligius for the last 3/4 days and I love the interface and the 0% fees, but isn't the variance extremely high?

I believe variance is high for slower minder due to the fact that min pdiff was increased from 1 to 2. At least that is how I explain my observation since it was bumped.

   T

I have a few TH/s and it looks to me variance is high

BTW, how do you cange min diff. to +256? It seems is not working on "my eligius" as I see cgminer mining at 127...

I see a bit of high variance on my end as well.

When you make your setting in "my eiligius" to 256  and save it- what happens? Does it remain 256 in the box? It should.

Yes, it stays... But at the same time it says "*** Minimum Difficulty Setting does NOT yet have any effect! ***"... Why is that?

[xyzzy099]

I've been mining at Eligius for the last 3/4 days and I love the interface and the 0% fees, but isn't the variance extremely high?

I believe variance is high for slower minder due to the fact that min pdiff was increased from 1 to 2. At least that is how I explain my observation since it was bumped.

   T

I have a few TH/s and it looks to me variance is high

BTW, how do you cange min diff. to +256? It seems is not working on "my eligius" as I see cgminer mining at 127...

I see a bit of high variance on my end as well.

When you make your setting in "my eiligius" to 256  and save it- what happens? Does it remain 256 in the box? It should.

Yes, it stays... But at the same time it says "*** Minimum Difficulty Setting does NOT yet have any effect! ***"... Why is that?

I remember seeing that before once when I had changed my dif.. I'm not sure if it has to wait until next round and then the settings kick in? - I'm sure someone can chime in

I'm pretty sure that functionality is just not implemented yet in the UI.  Eligius also ignores the '--request-diff' parameter from BFGMiner.

[Rampion]

Something funny is happening at Eligius. My 3 hour hashrate dropped by +20%, and the graph is showing vert funky behaviour for ALL my miners... I was terribly worried about something electrical going on in my farm (you know, hashrate fluctuating wildly in all the miners, at the same time) but I checked the graph of other users and it seems that its the pool who is having the problem... Obviously cgminer show the miners hashing just fine.