[~1000 GH/sec] BTC Guild - 0% Fee Pool, LP, SSL, Full Precision, and More

[phelix]

BTCguild was already my favorite pool, and I've kept cards pointed at it as best as I can even during the DDoS and other problems.

Now that Eleuthria has actually done something about pool hopping, I like the pool even more. Especially since it was really clear that the decision was made carefully and not just a knee-jerk reaction to one person complaining or the like...

Everyone who's complaining? Aww, too bad, there's one less pool they can rip off the legitimate users from. Ignore them, Eleuthria! You've done the right thing for the majority of dedicated users. Let the complainers take their intermittent contributions to another pool.

+1

soon all pools will have some sort of protection against hopping. for most people having to manage hopping thus making minig even more complicated is just a bothersome hassle.

[1714202806]

1714202806

[Dubs420]

Mmmh, 2 weeks ago, eleuthria said to me, he wont do anything against pool hopping, Congratz to changing your mind. But i think he will loose a lot of donations, but in the end it is healthier for the pool.

He gained a 2.5% donator right here.

[lemonginger]

I think Eleuthria has proven 100x over that he isn't just in this for the money and cares about the pool and the BTC community. There has been no pool that has so transparently and effectively dealt with exponential growth, DDOS attacks, etc.

I see no possible reason why having to wait an hour past finding a block to get paid would be problematic.

In any case, another +1000 to Eleuthria and BTCguild.

[jjiimm_64]

my 2 bitcents.

since the implementation of the delayed stats, my rewards have been scarily stable.  I get the same rewards for the really short blocks or the really long blocks. (~.13)

before the delayed stats, the very short blocks would range from .03 to .18 .  I can only assume the variance was caused by many miners hitting the pool at the beginning of the short block.  

I'm sticking with the guild, and keeping my donate at 5%.


edti:  elue:  any chance on adding the ability to set a local timezone for displaying times?  apparently I cannot subtract 5 hours in my head...  or is it 4 lol

[PcChip]

If anything those who are "butt hurt" are the ones who applaud this move by Urethra.

LOL


I agree his name is hard to spell, but...

http://en.wikipedia.org/wiki/Urethra

is slightly different than

http://www.eleuthria.com/image/site/logo.png

[hugolp]

my 2 bitcents.

since the implementation of the delayed stats, my rewards have been scarily stable.  I get the same rewards for the really short blocks or the really long blocks. (~.13)

before the delayed stats, the very short blocks would range from .03 to .18 .  I can only assume the variance was caused by many miners hitting the pool at the beginning of the short block.  

Same here. With the hoopers gone the payout is a lot more stable.

You can see the measure hurt the hoopers by the reactions in this thread.

I was thinking, and the pool hoopers could know sometimes when a round starts. When the previous round was longer than 1 hour the round will appear and the pool hoopers will know a new round has started. Its a bit of a long shot, but I think the solution is quite easy. It could be solved by delaying 10 minutes the report of a block that was longer than 1 hour.

I dont know what I was thinking when I wrote this. For some reason I though the dealy started counting when the round started...  

[eleuthria]

Added the ability to turn idle miner warning emails on/off for individual workers.  Mostly for those who are mixing CPUs with their GPU mining.

[fcmatt]

the only way to prevent hopping is by removing/closing the pool

why is closing/removing the pool the only way to stop pool hopping?

i asked this question to tycho over in his thread about deepbit. My question is underlined while
tycho's response is in bold.

tycho,

i have been reading that some people claim there is a way to use your pool as one of the pools for a pool hopping
strategy. specifically it was this website making a claim but naturally they demand a donation to get in on the "secret".

http://fasthoop.appspot.com/

based on my understanding of how your pool delays stats i am not quite figuring out how they can do this.
are they assuming if a block is solved and they cannot figure out who did it via screen scraping or an api.. it
must be deepbit?

or do you possibly know something i do not?

There are some possible options, but usually they don't give 100% true resuts.
The fact that they don't show it for everyone is a bit suspicious, by the way.

All pool hashrate jumps and dips are visible to me, so if poolhopping will cause a considerable loss for my users, I'll take appropriate actions.


based on that.. now that btcguild and deepbit both delay stats.. i am going to bet the possible options that
tycho mentions are even worse now percentage wise until i figure out otherwise. i have some ideas about this but i am still
reading and looking for hints which i think i found in another thread.

-------
oh. look at this. fasthoop.appspot.com just changed up their text on the website:

"Update: deepbit support is only provided to donators(only valid before btcguild's change). PM techwtf at forum.bitcoin.org for more details. The data may not accurate if pool is DDOSed.
Latest update: btcmine data is temporary unavailable"

it appears they were making an assumption if a block was solved it must have been deepbit. that is now broken as the
results were not as desirable as before.

now in the same thread about pool hopping the ideas they are coming up with are no where near as reliable.

[flower1024]

somebody said you could detect it through bitcoin itself.

just patch your bitcoin to do > 100 or more connections.
when a new block is announced have a look at the sender.

if the first announce came from btc guild its most likely their block.

its not my idea. but i am currently thinking about writing a patch for bitcoin.

[sharky112065]

the only way to prevent hopping is by removing/closing the pool

why is closing/removing the pool the only way to stop pool hopping?

i asked this question to tycho over in his thread about deepbit. My question is underlined while
tycho's response is in bold.

tycho,

i have been reading that some people claim there is a way to use your pool as one of the pools for a pool hopping
strategy. specifically it was this website making a claim but naturally they demand a donation to get in on the "secret".

http://fasthoop.appspot.com/

based on my understanding of how your pool delays stats i am not quite figuring out how they can do this.
are they assuming if a block is solved and they cannot figure out who did it via screen scraping or an api.. it
must be deepbit?

or do you possibly know something i do not?

There are some possible options, but usually they don't give 100% true resuts.
The fact that they don't show it for everyone is a bit suspicious, by the way.

All pool hashrate jumps and dips are visible to me, so if poolhopping will cause a considerable loss for my users, I'll take appropriate actions.


based on that.. now that btcguild and deepbit both delay stats.. i am going to bet the possible options that
tycho mentions are even worse now percentage wise until i figure out otherwise. i have some ideas about this but i am still
reading and looking for hints which i think i found in another thread.

They have scratched out their donation link. Guess they were using other pools and BTC Guild to determine when to  hop on Deepbit.

http://fasthoop.appspot.com/

Update: deepbit support is only provided to donators(only valid before btcguild's change).

[ellipsis]

its not my idea. but i am currently thinking about writing a patch for bitcoin.

The hub mode patch is probably sufficient.

[fcmatt]

somebody said you could detect it through bitcoin itself.

just patch your bitcoin to do > 100 or more connections.
when a new block is announced have a look at the sender.

if the first announce came from btc guild its most likely their block.

its not my idea. but i am currently thinking about writing a patch for bitcoin.

yup. i was just reading about that. at first glance it does seem like it could work.
in practice i am not sure. one is talking about milliseconds being the determining factor
in deciding when to switch.

[flower1024]

yup. i was just reading about that. at first glance it does seem like it could work.
in practice i am not sure. one is talking about milliseconds being the determining factor
in deciding when to switch.

if it were ms i KNOW it will work
i just have to include ping times...

[fcmatt]

yup. i was just reading about that. at first glance it does seem like it could work.
in practice i am not sure. one is talking about milliseconds being the determining factor
in deciding when to switch.

if it were ms i KNOW it will work
i just have to include ping times...

 couldn't a pool operator just allow 25-50 (more?) trusted IPs to connect to his bitcoin
daemon and stop this attack cold? yes.. they need to broadcast the found block ASAP
but they could just say it to 50 nodes who they know who they are, geographically located
in strategic places, etc... i am sure digging up a list is not too hard especially if pool operators
cooperated a bit..

and i really do not know how ping (icmp) will compare to a bitcoin client saying i found a block...
there might be a lot of variance in that considering icmp is given pretty low priority on most backbone
routers and what not... and a pool operator could just not allow it period. so you will end up pinging
their default gateway at the ISP or a machine close to them in the same subnet perhaps.

[flower1024]

as long as their blocks have a fixed delay from a pre known node it shouldn't be a problem.

i don't know if the pool operators could live with only 50 nodes. as far as i know their goal is to spread new blocks as fast and wide as possible (this counts especially for btcguild as elu pays for invalid blocks and looses money if he has an high invalid rate)

i just try.
i will publish my experience to everyony.

my goal is not to pool hop my self. i want to stop it in the long term. but if we want to stop it we have to understand it completely.

i don't like the idea, that a few "high-minded" people are able to cheat the rest.

everybody could cheat; or nobody: thats what i am trying to archive

[ptshamrock]

Some of my miners always become idle with btc guild.. why is this?? Phoenix 1,5 vectors bf_int agression 11

[fcmatt]

as long as their blocks have a fixed delay from a pre known node it shouldn't be a problem.

i don't know if the pool operators could live with only 50 nodes. as far as i know their goal is to spread new blocks as fast and wide as possible (this counts especially for btcguild as elu pays for invalid blocks and looses money if he has an high invalid rate)

i just try.
i will publish my experience to everyony.

my goal is not to pool hop my self. i want to stop it in the long term. but if we want to stop it we have to understand it completely.

i don't like the idea, that a few "high-minded" people are able to cheat the rest.

everybody could cheat; or nobody: thats what i am trying to archive

you know... it is really hard to take you seriously about what you are saying because you pool
hop as i type this based on your other posts to other threads.

if your plan/proof of concept code works.. i would expect you to start using it immediately.. to pool hop.

but that is just me.. i am not very trusting eh?

[eleuthria]

Some of my miners always become idle with btc guild.. why is this?? Phoenix 1,5 vectors bf_int agression 11

Phoenix seems to occasionally idle for a few seconds after an LP.  My guess is you get your LP work, and you're lucky enough to find a solution almost instantly and send it in.  Because the LP is very resource/network intensive, the server's responses are a bit delayed.

I've noticed the new poclbm has not had a single idle, while phoenix has them once in a while at the LPs.   Phoenix seems to discard a getwork immediately after finding a share, while poclbm continues and occasionally finds multiple shares in one getwork space.  This not only means poclbm doesn't demand new work as often (helping the server), it also means that it doesn't fire off for new work immediately after an LP if it finds a share early in the getwork.

[ptshamrock]

So what do you recommend now? How to use poclbm? I have 5870 at 950/300 400mhash each 36 of them..

[fcmatt]

So what do you recommend now? How to use poclbm? I have 5870 at 950/300 400mhash each 36 of them..

i guess sharing my results of using phoenix versus poclbm is that i am not seeing any real difference.
pocblm is currently giving me .07% stales while phoenix is a touch above that being around .08-.09%.
i am on win7 x64 using guiminer and pheonix is using the phatk kernel.