Quote from: dexX7 on March 24, 2014, 08:34:20 AM

Hey guys,

so to my understanding the reason for the idea of making multisig transactions non-standard was the high amount of unspent outputs that were created by data-storage transactions. I want to put this into some perspective:

(http://i.imgur.com/c9mpd8h.png)

I did not evaluate how many unspent outputs were created by all multisig outputs, because I'm unable to do so efficiently, but here are some observations nevertheless:

Out of 35.508.561 multisig outputs ever seen on the blockchain XCP and MSC account for 5.192 multisig outputs. That's about 0.000146 %.
Out of 9.819.223 unspent outputs at the very moment (RPC: gettxoutsetinfo) the amount of unspent XCP and MSC multisig outputs is 3.226. That's about 0.000329 %.

In the case one wants to run some tests on this data: multisig-usage.rar (format: "txid vout rawtxhex" -- file size is about 8 MB, but be aware that multisig-all-full.txt is over 1.4 GB unpacked!)

Please don't get me wrong, I do not say that we don't contribute to a higher utxo set and a general statement about the usage of multisig transactions is more difficult to make, because I have no idea how to evaluate the other 99.999%. My above used criteria were: "has an standard output to Exodus address and multisig output" and "has an multisig output with XCP magic bytes". If you have any idea, go ahead.

Thank you for the intelligent analysis dexX7. The numbers are pretty stark when you "see" them.