bitHopper: Python Pool Hopper Proxy

[c00w]

Um, I just redid the whole deepbit system again to produce estimated share counts. So every scheduler now works!

[simonk83]

@simonk83 happened to me a moment ago with mt.red , just deleted the log thinking it's a connection prob with them, seems like it happens with a server it chosen by bH, please run it with

Code:

--debug > output.log

if you dont mind

I realised I wasn't quite on the latest, so I've done a pull.  If it happens again I'll run it via debug and grab the log.  Seems OK for now though (although it was running for a few hours prior to those errors happening).

[joulesbeef]

Um, I just redid the whole deepbit system again to produce estimated share counts. So every scheduler now works!

kick butt c00w


and i just submitted my first deepbit share.. WOOHOO
lol
one the one so far

but I havent upgraded yet. i'm on 1.3-63

[joulesbeef]

Ok who erased OldDefaultScheduler from the wiki?

it's a valid scheduler..which some think might be the best one there is. Just cause it is called old, no need to erase it from existence.

[c00w]

Wiki history shows who did what. I'll track em down

[paraipan]

Ok who erased OldDefaultScheduler from the wiki?

it's a valid scheduler..which some think might be the best one there is. Just cause it is called old, no need to erase it from existence.

me did it  , sorry

[joulesbeef]

It was paraipanakos!! I knew it! He always had it in for OldDefaultScheduler.

You should code into the next hopper, If user = paraipanakos then mine+charity everything  

[GenTarkin]

LOL! So, question what exactly is the diff between the old default scheduler and the new one. I notice the old one doesnt list the "slice" column...

[paraipan]

It was paraipanakos!! I knew it! He always had it in for OldDefaultScheduler.

You should code into the next hopper, If user = paraipanakos then mine+charity everything  

  nice, I was asking for it

edit: (mental note) don't mess with things I can't remember what are used for

[djex]

Ok who erased OldDefaultScheduler from the wiki?

it's a valid scheduler..which some think might be the best one there is. Just cause it is called old, no need to erase it from existence.

IMO best one regards to payout. I found with slicing my payout is often below the daily expected payout for my hash rate but with the OldDefaultScheduler payout is around 0.28 BTC more than the expected.

[bb]

@GenTarkin: old one doesn't slice, it just (more or less) stays at the pool with the lowest share count.

@djex: slicing has only been around for so long. Give it some time to kill the variance.

[djex]

So with mine_deepbit I should switch to info on all possible pools (out of the ones disabled) that will give proper stats?

[muyoso]

Can't wait to hear the first reports back on the deepbit hopping.  Been excited ever since I read about it yesterday.  

[joulesbeef]

LOL! So, question what exactly is the diff between the old default scheduler and the new one. I notice the old one doesnt list the "slice" column...

see this discussion here


i'm giving him a hard time with extremes, but there is a lot of evidence to support him that it is the best scheduler

I'm not 100% sold due to some slow pools, but I'm running it now. Have to give due warning about the stats page.. it's different

which brings up.. who is most responsible for the altslicescheduler?

I just have some questions on how it works, for organofcorti to sim.

So with mine_deepbit I should switch to info on all possible pools (out of the ones disabled) that will give proper stats?

yes
i add BTCmine as well.. it was never hoppable but it is one of the ten largest pools.
not sure if it is in the hopper.

[c00w]

@djex

Um we actually don't use the other servers in the latest iteration. We probably should. But our cutoff times are a lot better.

[Clipse]

I recon there is an extra method to "slicing" that wasnt considered here yet.

Adjusting mining method based on time left till next difficulty and considering current difficulty to next difficulty increase/decrease.

If next difficulty is soon (insert random low timeframe) and difficulty is increasing (insert random increase percentage here) then stick to (insert random fast pools with lower share)
If next difficulty is soon (insert random low timeframe) and difficulty is decreasing (insert random decrease percentage here) then ignore (insert random pool speed and only consider lowest share)

If next difficulty is in a while (insert random high timeframe) and difficulty is increasing (insert random increase percentage here) then ignore (insert random pool speed and only consider min shares)


I think you could even narrow the timeframes down with some neat formula considering pool avg daily hashrate , share count , next difficulty(up or down) and lastly timeframe till next difficulty adjustment.

Its important to consider difficulty adjustments when mining low share slow pools since they can easily spread across difficulties where you will get penalised even more.

This approach should make the whole lowest shares vs fastest pool debate move more smoothly.

[joulesbeef]

@djex

Um we actually don't use the other servers in the latest iteration. We probably should. But our cutoff times are a lot better.

oh so I dont have to have 20 info servers huh..cool that will make my screen smaller again.

what about bitcoin.lc.. does leaving them on hurt? they do that random delay crap

[GenTarkin]

So, I just saw a bunch of new blocks pop up like 2 or 3 ... but, BH said they were owned by 2 diff pools in the LP output....but those 2 pools on the stats page did not get their shares reset... Im confused...shouldnt the shares reset if they found new blocks? (if the LP implementation is working properly in BH?)

[Clipse]

So, I just saw a bunch of new blocks pop up like 2 or 3 ... but, BH said they were owned by 2 diff pools in the LP output....but those 2 pools on the stats page did not get their shares reset... Im confused...shouldnt the shares reset if they found new blocks? (if the LP implementation is working properly in BH?)

Depends which pools (pools in question could be delaying stats) or their blocks were invalid for some reason and they simply continue counting.

[joulesbeef]

gah I'm torn between oldDefaultScheduler and DefaultScheduler
BTCServ.net is new and slow.. with old default I am stuck here until someone finds a block...crap i dont even know where to find balance or crap on that site.
to slice or to not slice that is the question.