Lol. Busted.

Ok so for somebody to be lucky enough to steal the bitcoins out of my main backup address...

They first need to generate my address which could be any one of 2^160 possibilities. And then they must generate the corresponding public key that gives them access to the bitcoins in my address? Of where there at 2^96 possibilities?

Or must they simple generate 2^96 bitcoin keys... at which point they would control every single bitcoin address in the world?

If you don't send your coins to/from an IP obviously

Say you generated a block of 50 coins (ie. didn't buy them anywhere, no trail yet) and then spent them from your machine later to purchase something you didn't want anybody to know about. Or the funds just happen to go to an unscrupulous individual in general and when he cashes them out them somebody wants to speak to you (the person who funded him) for some reason or another let's say.

Could somebody tell simply by looking at the block chain or other factors what ip address sent or generated what coins?

I know messages can be sent with blocks. Is any information related to the owner's computer stored at all in the block chain basically is my question?

This just happened in effect to the namecoin blockchain.

Wikileaks boasted about it's greatness and overnight the difficulty jumped from a few thousand to 55,000.

Now the community is struggling to solve a single block every few hours and the difficulty is going to go back down eventually, but every day that "next difficulty change" time jumps another few days... so it's going to take a while!

So yes it is possible, but it will not be possible once bitcoin really goes mainstream, and that is all that matters.

Ok back with an update.

I tried running both poclbm and phoenix miners on multiple different video cards against bitcoind on port 8332 running testnet.

Both miners showed considerable "miner is idle" messages in between solved blocks for a number of hours.

I also have pushpool running on the same server.

I just switched one of the miners over the port 8344 and am mining shares through pushpool now (all other settings are the same) and thankfully pushpool is doing it's job flawlessly. A few hundred shares found so far with no "miner is idle" messages.

So I guess bitcoind just doesn't function exactly as expected when you are controlling 75%+ of the network... who'd a thunk?!

How can you be so sure it is not a power problem?

How much power do you have dedicated to your 6990s? What PSU are you using exactly?

The fact that the problems start whenever the cards go over a certain use percentage almost tells me exactly that power is the real source of the problem. Your cards draw more electricity as they ramp up in terms of "usage percentage". ie when your card is using 100% usage you are probably pulling ~300 watts each and when you are down to 50% usage you could only be doing 150 watts.

Nope. Never seen that. And yes, I was able to install pushpool on the latest ubuntu 11.04 server edition.

QUESTION: is /tmp/shares.log and /tmp/request.log entirely necessary? Under what circumstances would one require those files? I assume they would become quite massive in a short amount of time.

I was thinking of simply backing up the shares mysql table every day in case I needed to do a rollback or something perhaps. Can't really imagine why I would ever need to do one of those either, however.

As soon as a block is found you may as well split up all the earnings immedately and EMPTY the MySQL Shares table, no? Or perhaps constantly empty all shares that are over 7 days old..? Or 2 blocks behind?

If no pool owners will, then perhaps somebody who knows more about pushpool and/or bitcoin in general can help me out?

I have proposed a solution to ddosing the bitcoind client's port on this thread.

Your hash-rate sounds about right.

Anybody posting the "extreme" hash rates is most definitely using water cooling. You cannot achieve those rates without water cooling.

Can I make a suggestion?

Save you money!

Buy another card instead of a water block if anything. They're almost the same price.

Also you would be capable of overclocking the 5770 10% without any cooling problems.

If it runs fine for mining, it will run much much cooler for gaming.

I have 23 cards all air cooled running for months straight now and there has yet to be a problem.

I can only think of one...

Choose a list of ~100 ip address that are constantly "connectable" found by tracking this graph every week or two: http://maps.google.com/maps?q=http://c1958612.r12.cf0.rackcdn.com/bitcoin.kml (not actually that graph, because it is a backup, apparently the live version by MagicalTux is down right now)

and add them to your firewall? To only allow those IPs to connect to port 8332 or whatever port you were running bitcoind on?

Another option would possibly be changing (to say port 52395 or something) and hiding your bitcoind port, creating the same firewall mentioned above, and marking it "invisible" to any other "port scanners" or what not? That way nobody would even know it existed or that the server was online (unless they were one of the 100 people you automatically connected to).

Ideas? Thoughts?

How many possibly bitcoin addresses are there exactly?

How long does it take to generate one?

If this is true...


Emphasis mine.

Then why do we not simply force the use of the windows security center into all windows builds of bitcoin by default (or enforce the windows security center to be running unless the user is "advanced" and clicks otherwise)...

The windows security center for those are are not aware is a notification nag system on windows that constantly nags the user to download the latest virus scanner, make sure their firewall is up, to enable the virus scanner and all updates, etc...

If every "basic" windows user was forced to do this by default, I think we could cut down on 90% of the possible thefts by trojan at least - all without much work on our part.

Linux systems would be considered safe for now from trojans and would not require the default security settings.

Like every time I mine a block almost, I am told that the miner is idle.

Could it be that since I am personally mining about 16 blocks per hour that the small bitcoin testnet distribution is not verifying these transactions fast enough?

I have just begun to run phoenix miner in an attempt to see if it is just the miner... however I have run both miners on the regular blockchain recently and have not received this number of "job finished" errors...

There appear to be smooth runs of poclbm miner that show next to no "miner is idle" warnings for quite a few blocks in a row, and then all miners seem to go to spotty chunks of mining time where every other block found there is a "miner is idle" warning.

Could it also be that I am maxing out my "m1.large" EC2 instance in terms of power? (I highly doubt it however, as it is showing between 99% and 100% "idle" CPU with all other values at or near 0% in the linux top command.

I am showing 27 connections on testnet. My port is forwarded.

There are approximately 10 5870s on 2 machines transmitting about 3,7Ghps from my home to the amazon cloud.

I will report back with the results from phoenix miner.

Hehe, thank you for the suggestions.

Thankfully my OCD isn't THAT out of whack

Take care,

I am working on creating a mysql database and don't want to waste any extra storage space.

So what should I make an entry for a bitcoin address?

CHAR X length? (is the address consistently one length?)

Or VARCHAR X length? (does the address change in length from user to user?)

Which installation are you referring to exactly? What release version?

So you only ever played with the flags? Not the BIOS and/or clock settings?

What have you changed? Can you remember?

Awesome, thank you for sharing!

No, actually.

I have found a way to circumvent the difficulty.

I can still mine at a difficulty of 1 if I want to, but I don't.