[4+ EH] Slush Pool (slushpool.com); Overt AsicBoost; World First Mining Pool

[slush]

i just activated my account but the website is telling me user/pass credentials don't match. can you perhaps send me a password reset or reminder?
i think this is weird as i just registered. i usually remember my passwords at least for a few minutes...

I tested registration few times today and I didn't get any troubles here. I deleted your account, please register again...

[1715013566]

1715013566

[1715013566]

1715013566

[1715013566]

1715013566

[slush]

I'm going to move application to new server now. There will be some service outage.

[chorchonga]

Thanks for the heads up!  Just made my first .001 bitcoin!

[slush]

I just finished migration of server. Old IP should expire in one hour, then I will shut down reverse proxy on old server, so old IP will not work. If you are using host name 'mining.bitcoin.cz', change is transparent for you and you don't need to change anything on your side.

[j16sdiz]

Some feature request ...

 - Sorting the worker stat  by the worker name

 - batch create workers (or  after create the worker, go back to profile page )


Why? Because I am trying to build a big cluster of workers (~30 hosts)..... (I may will give more details later)

[BitterTea]

Somebody joined the party... 12 ghash/sec now...

[fabianhjr]

LOL this is really growing fast. Slush you have 1/10 of the network in your pool. How does it feel?

[hacim]

well, I got my 1btc finally... after three weeks of collaborative mining. I think I'm out, doesn't seem worth it to burn my cpus that hard.

[ledskof]

well, I got my 1btc finally... after three weeks of collaborative mining. I think I'm out, doesn't seem worth it to burn my cpus that hard.

It's not really worth it if you're using actual CPUs instead of GPUs.

[slush]

- Sorting the worker stat  by the worker name

Done on dev.

- batch create workers (or  after create the worker, go back to profile page )

Actually I'm thinking about make it _harder_ to create botnets. Adding captcha for user and worker registration and limit worker count to something like 10 per account, not allow frequent IP changes of workers and so. Connecting botnet of poor workers to pool is my nightmare for both performance and legal troubles.

Why? Because I am trying to build a big cluster of workers (~30 hosts)..... (I may will give more details later)

I made pool to help standalone miners smooth their income, not to help people with stealing CPU power from university computer lab.

[slush]

well, I got my 1btc finally... after three weeks of collaborative mining. I think I'm out, doesn't seem worth it to burn my cpus that hard.

I think pool is useful to get first newcomer's Bitcoin quite easily even when he has only poor CPU. But it is definitely economic nonsense to mine with CPU to make money. So congrats for you first mined bitcoin, but I believe you can obtain more bitcoins much cheaper on #bitcoin-otc IRC channel :-).

[BitterTea]

well, I got my 1btc finally... after three weeks of collaborative mining. I think I'm out, doesn't seem worth it to burn my cpus that hard.

It's not really worth it if you're using actual CPUs instead of GPUs.

It's barely worth it, probably even not, with a GPU miner as well. The interesting thing about Bitcoin, is that one could make the case that people are too focused on mining, but the act of contributing that processing power helps the network become stronger. Blows my mind sometimes.

[fabianhjr]

You might want to update the hashrate because we are near 14GHashes/sec now.

Also, my miner found 2 blocks already. YEAH!

[adv]

limit worker count to something like 10 per account, not allow frequent IP changes of workers and so.

What could prevent man-with-class-room to use the stupid proxy? (or not so stupid: http://bitcointalk.org/index.php?topic=2431.0)

What is the difference (for performance) is plugged into a worker one powerful or many weak miners?

[slush]

What could prevent man-with-class-room to use the stupid proxy? (or not so stupid: http://bitcointalk.org/index.php?topic=2431.0)

I think IP changes are prevention for botnets than for class rooms. If you use proxy, you have additional network overhead and maybe you lost anonymity because your proxy have to run somewhere. It's little harder (for script kiddie) to setup fast anonymous proxy correctly. I agree it is mainly security by obscurity, but this should not affect regular users with two three miners...

What is the difference (for performance) is plugged into a worker one powerful or many weak miners?

The difference is extreme. Miners need fresh getwork every few seconds to crunching actual block. So typically miners download job every 5-6 seconds. From current pool stats you see, that getwork rate is at 85 getworks per second. Diablo miner for ATI 5970 download ~6 getworks per 5 seconds for 600mhash/s (so approx. 1 getwork per second), but 300khash/s CPU miner call getwork once per 5 seconds (0.2 getwork per second).

Situation 1) All miners are using ATI 5970. For 85 getworks/s the pool performance is 85*600  = 51000 mhash per second
Situation 2) All miners are using CPUs @ 300khash/s, the pool performance is 85/0.2*0.3 =  128 mhash per second

Both for the same server load. See the difference?

Another view can be total getworks per found block. When miner have rate 1 getwork/s and average time for block is 27 hours, it is 97000 getworks/block. When CPU miner has average time 1 year, you need 6300000 getworks/block. You spend much more resources on network overhead and server capacity than 50 BTCs have current value (not counting power needed for miner itself). So mining with such poor CPU miners is completely worthless. Supporting CPU miners in pool is charity from all GPU donators, because pool with CPU miners will never pay itself for server hosting. I'm happy for CPU pools miners crunching blocks on their personal netbooks for their first bitcoin, but I definitely don't want botnet of 10year old crap computers from computer labs.

[adv]

I can not understand why getwork/s does not strictly proportional to the power of the miner. I think in theory it should be: 100 * (1M h/s : 0.01 gw/s) == 100M h/s : 1 gw/s

In practice, for your own statistics, it turns out that many weak miners more often do a getwork request, it means that they solve a larger number of blocks, is not it?

[slush]

In practice, for your own statistics, it turns out that many weak miners more often do a getwork request, it means that they solve a larger number of blocks, is not it?

Because slow miners are not crunching whole 32bit nonce space. They must ask more frequently to work on up-to-date jobs. For example, jgarzik's miner is iterating only over 24 bits from nonce and GPU miners are using 32bits, but they also have x second timeouts.

[slush]

I just enabled automatic pool donations and email notifications on payouts. Reporting worker crashes will be soon.

Also thanks to all who set up some donation on profile page :-).

[fabianhjr]

I set up a 6% donation cus you are awesome!

[Raulo]

From what value is this donation is calculated? From unconfirmed reward, confirmed reward or reward sent? I've just got a bunch of rewards and nothing was subtracted.