What is the purpose of this transaction?

[hhanh00]

https://blockchain.info/tx/ee61292f8a95584b73bce028e91df2c93ec11fa3636ac65c8fa697fcb6e2c3a1

It gathers a bunch of strange tx outputs without actually moving anything. Is it to fix a bug in the mining pool software?

Thanks,
--h

[yakuza699]

One more question on top of this one: how did he created so many empty inputs?

[Muhammed Zakir]

I checked the tx, instead of showing the address it is showing non-standard hash. Maybe the addresses are invalid or maybe there was some mistake when he signed it? But if there is any problem, it can't be pushed, right?

   ~~MZ~~

[chanz]

Can't see no purpose of this their address have many similar transactions like this. They must be testing something or their software screwed big time. But that "Unable to decode input address" just means that the transaction does not fit the usual address metaphor or atleast that's how blockchain sees it.

[shorena]

If you use blocktrail instead of bc.i you can follow the inputs. https://www.blocktrail.com/tx/ee61292f8a95584b73bce028e91df2c93ec11fa3636ac65c8fa697fcb6e2c3a1

Looks they are from inputs that came from a new block. Not sure how a miner would do this.

[laurentmt]

I think it's related to P2Pool. It seems that when it founds a new bitcoin block, P2Pool adds a "marker" in the coinbase transaction (see last output from this tx).

This output script (0x20 0x83e03b27ad09e5128d13a84c1154bba2a526f65f52a6d4e6759a8edfe4419568) says push 32 bytes (0x20) of data (0x83e03...) onto the stack when you must evaluate the script. This is a very strange output script but I guess it's related to how P2Pool works.

The transaction you've spotted cleans the UTXO set by "spending" these dust outputs and my best guess is that the address "1HfA1KHC7bT1hnPPCjpj9CB4koLM4Hz8Va" is controlled by P2Pool.

[hhanh00]

It happens hundreds of times. I was wondering if it was on purpose and if so, why?

[gmaxwell]

P2pool used to store its sharechain commitments (which is how it achieves trustless, serverless pooled mining) in a zero value anyone can spend output.  To clean up the utxo set some people spent all those outputs at some point. (that might have even been me making that txn, dunno).

It now uses a op_return... now that bitcoin core knows it doesn't need to track those, so there is no useless utxo to clean up.

[hhanh00]

Thanks - that makes perfect sense.