Who need free bitcoins?

[pent]

Just come and get some

http://blockexplorer.com/tx/41836560e2439f440514af96ca394a38bad6f3d9d0d11dba667c886b16e504ec

Code:

0.00224733	Unknown	Strange	OP_IFDUP OP_IF OP_2SWAP OP_VERIFY OP_2OVER OP_DEPTH

Output does not contain CHECKSIG's!

I tried to craft needed transaction input, but after several fail attempts my client has been banned eveywhere. :lol:

[pent]

Hint: This is not only one such transaction in chain.

Seems that p2pool produces them. Ammm... WHY?

[adamstgBit]

..
http://blockexplorer.com/tx/41836560e2439f440514af96ca394a38bad6f3d9d0d11dba667c886b16e504ec

Code:

0.00224733	[u]Unknown	Strange[/u]	OP_IFDUP OP_IF OP_2SWAP OP_VERIFY OP_2OVER OP_DEPTH

...

whats going on here?

[pent]

..
http://blockexplorer.com/tx/41836560e2439f440514af96ca394a38bad6f3d9d0d11dba667c886b16e504ec

Code:

0.00224733	[u]Unknown	Strange[/u]	OP_IFDUP OP_IF OP_2SWAP OP_VERIFY OP_2OVER OP_DEPTH

...

whats going on here?

They could be redeemed without any keys and signatures.

[Fronnk]

..
http://blockexplorer.com/tx/41836560e2439f440514af96ca394a38bad6f3d9d0d11dba667c886b16e504ec

Code:

0.00224733	[u]Unknown	Strange[/u]	OP_IFDUP OP_IF OP_2SWAP OP_VERIFY OP_2OVER OP_DEPTH

...

whats going on here?

They could be redeemed without any keys and signatures.

So, how does one redeem them?

[Raoul Duke]

..
http://blockexplorer.com/tx/41836560e2439f440514af96ca394a38bad6f3d9d0d11dba667c886b16e504ec

Code:

0.00224733	[u]Unknown	Strange[/u]	OP_IFDUP OP_IF OP_2SWAP OP_VERIFY OP_2OVER OP_DEPTH

...

whats going on here?

They could be redeemed without any keys and signatures.

So, how does one redeem them?

Would like an answer to that also

[viboracecata]

What this mean? Who can get the lost bitcoins?

[Steve]

I looked at the script and it's not obvious to me what it might have been trying to do.  I'm guessing it's a bug of some sort in p2pool.

[piuk]

This script actually spells "script"

Op Code	Decimal	Ascii Code
OP_IFDUP	115	s
OP_IF	99	c
OP_2SWAP	114	r
OP_VERIFY	105	i
OP_2OVER	112	p
OP_DEPTH	116	t

http://blockchain.info/tx/41836560e2439f440514af96ca394a38bad6f3d9d0d11dba667c886b16e504ec?show_adv=true

[Steve]

This script actually spells "script"

Op Code	Decimal	Ascii Code
OP_IFDUP	115	s
OP_IF	99	c
OP_2SWAP	114	r
OP_VERIFY	105	i
OP_2OVER	112	p
OP_DEPTH	116	t

http://blockchain.info/tx/41836560e2439f440514af96ca394a38bad6f3d9d0d11dba667c886b16e504ec?show_adv=true

So I'm guessing a coding bug where the string "script" was put into the transaction instead of the contents of the variable named "script" …or perhaps it's an unfinished part of the code where they were just sticking "script" into the transaction as a placeholder.

[Gavin Andresen]

Those 0.002 bitcoins cannot be redeemed, because there is no input that will make that script valid (because there are no OP_ENDIF opcodes in there to match the OP_IFDUP/OP_IF opcodes).

[rjk]

Those 0.002 bitcoins cannot be redeemed, because there is no input that will make that script valid (because there are no OP_ENDIF opcodes in there to match the OP_IFDUP/OP_IF opcodes).

So is it actually possible to create transactions in the blockchain that don't go anywhere, until (first come first served) someone grabs them? That would be a fun easter-egg hunt/puzzle challenge for the super-nerds.

[pent]

Yeah, output can be redeemed by conditions provided in script. I.e. protected by password only or not protected at all.

[forrestv]

This is a bug in P2Pool. I had been assuming that this was someone mining to that script just for fun, since previous versions of the P2Pool protocol gave you the freedom to choose your payout script. Since then, the rules have tightened, and this should have been impossible.

P2Pool computes payouts using a PPLNS scheme, which entails counting the last share only partially, which requires a special case. The special case was broken - The fault was using Python's dict(script=...) syntax instead of {script: ...}. See http://pastebin.com/0RfRWZ21

Fixing this will break compatibility with current versions of P2Pool, so it will take some time.

Thanks for pointing this out.

[finway]

I should make a bot watching&grabbing these free coins. 

[Red Emerald]

This is a bug in P2Pool. I had been assuming that this was someone mining to that script just for fun, since previous versions of the P2Pool protocol gave you the freedom to choose your payout script. Since then, the rules have tightened, and this should have been impossible.

P2Pool computes payouts using a PPLNS scheme, which entails counting the last share only partially, which requires a special case. The special case was broken - The fault was using Python's dict(script=...) syntax instead of {script: ...}. See http://pastebin.com/0RfRWZ21

Fixing this will break compatibility with current versions of P2Pool, so it will take some time.

Thanks for pointing this out.

So who were those coins supposed to go to?

[forrestv]

So who were those coins supposed to go to?

It's the first share of the sharechain that was supposed to be paid out, which is supposed to be partially counted. It's incorrectly directed to a nonsense address, and so it takes from whoever had the last share in the payout window... translating to everyone in a small way, in the long run.

[dooglus]

I should make a bot watching&grabbing these free coins. 

1) They're not grabbable
2) In total they're only worth 0.38174231 BTC

[dooglus]

Those 0.002 bitcoins cannot be redeemed, because there is no input that will make that script valid (because there are no OP_ENDIF opcodes in there to match the OP_IFDUP/OP_IF opcodes).

OP_IFDUP doesn't need a matching OP_ENDIF opcode, but OP_IF does.

The end of EvalScript() checks that all OP_IF and OP_NOTIF operations have been closed by an OP_ENDIF, else returns false.

[dooglus]

P2Pool computes payouts using a PPLNS scheme, which entails counting the last share only partially, which requires a special case. The special case was broken - The fault was using Python's dict(script=...) syntax instead of {script: ...}. See http://pastebin.com/0RfRWZ21

Every p2pool block since 12th Feb 2012 has had this problem, with the exception of http://blockexplorer.com/b/169625, on 4th Mar 2012.  Somehow that one's OK.  That's 133 bad blocks and one good block.

What was different about that one block?