Bitcoin Forum

Open thread to discuss unusual or confusing transactions observed on http://blockexplorer.com , and possible explanations for them.

To start things off... does anyone know what's going on with this block? http://blockexplorer.com/b/115060

There are more than 70 transactions of over 1000 BTC in a single block!  Each transaction appears to feed into the next one, which is why they all appear in descending order of value.

Looks like Tycho's pool payout... For example, my pool's payouts looks like this (http://blockexplorer.com/block/000000000000af2f0072db102f7b6e53f071357d33e69d53e509e47dae424062)

(just adding this thread to my watchlist)

I guess most of the weird patterns I've observed would probably be pool payouts and such then.  Kind of a hit for anonymity.  Is the 1000 BTC that remains probably Tycho's profit, then?

I don't think so. Nobody really know who owns those addresses. And while mining standalone, anybody can watch your mining income, too, because once you spend some coins, they are connected together in the spending transactions.

If you care about anonymity, you can use pools over Tor as some people do.


Probably not, the rest on the account are payouts for people which don't cross their withdrawal limits.

From bitcoinmonitor.com, every 10 on the 10:

http://i30.photobucket.com/albums/c347/pi_factorial/bitcoinmonitor.png

Notice that every ten minutes on the dot there is a transaction of around 200 BTC.  A several of them are exactly 191.12 BTC.  For example:
http://blockexplorer.com/t/7JNH6MUkCn
http://blockexplorer.com/t/4B52WFCPx2

And yet, the breakdown of the 119.12 is completely different.

Any ideas?

it is, I see my wallet listed there :)

It took 12 minutes to calculate this block: http://blockexplorer.com/b/117147

There were more than 90 unconfirmed transaction at the time the block before this finished, and at the time this one finished.  Why did only one (non-generation) transaction appear in the block?

a) Somebody is accepting only transactions with fees (edit: most probable; there is currently only one more transaction with fees (http://blockexplorer.com/search/bc8a5f9007bb55cdb5b27477fe86cc4dc5b5ee00e9817ff5037512351c426860), which is 'sendmany' payout from the pool. Sendmany transactions are accepted only by newest clients, so older clients ignore this (their bad, they are missing my fees just because they use older bitcoin version :).

or

b) Somebody started bitcoin client and found block pretty quick. Older transactions are not actively relayed to freshly connected node, so it's pretty normal that bitcoin client does not know all pending transactions.

Recently i saw some nice pattern - ~8+ blocks found in pairs constantly in the day before difficulty step :)

Discovered via Bitcoin Monitor:

Block found at 2011-04-11 18:52:57
http://blockexplorer.com/a/2EnJ1VL8Kx

Bitcoin Charts reports the following trade on MtGox:
Apr 11, 2011, 18:53:44   |   0.75618   |   50.00   

Looks like some miner's got a bot to sell off block bounties.  This is the first time I've seen this particular pattern.

http://blockexplorer.com/b/118777

Total BTC: 95069.09

You don't see that every day.

Four transactions over 10000 BTC:
http://blockexplorer.com/t/2s9dR7tzdg
http://blockexplorer.com/t/6CoNTeXvvj
http://blockexplorer.com/t/3SMDDjhy4V
http://blockexplorer.com/t/2wcj9do8EA

Block discovery bounties can be paid directly to multiple addresses?
http://blockexplorer.com/b/121668

Yep. That's Luke-jr's pool. Puddinpop's pool also used to do that. It's nice because all participants get paid immediately instead of having to wait 120 blocks.

Well, that's not absolutely correct. People _see_ block payments immediately, but also have to wait 120 blocks till it mature (or 100 blocks with hacked bitcoin client).

As an example, here's one of the blocks generated with Puddinpop's pool - http://blockexplorer.com/b/99004

Oh, I get it - when the client generates the block, it includes the transaction for the generation bounty in the block, so that transaction can pay out to anyone, and this is valid so long as the total payout = 50 BTC (or whatever the bounty is at the time) + transaction fees.  Right?

Would it be possible for a miner to create extremely erratic payouts (i.e. distribute the bounty to a million addresses) as a vector of attack?  Or would the block be rejected in that particular case for being over 1MB?

Right. The generation is just an input to a transaction. The value of the input is found by the block subsidy + fees. The outputs follow the normal transaction rules.

The generator could easily create a giant generation transaction with tons of garbage outputs. You don't even need to pay out to the outputs: 0-value outputs are valid. The block size is limited to 1MB, though.