rpc: consolidatemsunspent changes.

[iFoggz]

Refactor, change explanation, go in more depth and use bytes instead since hex is more confusing.

Gone more in depth and researched almost every aspect to give better explanation to the curious where the numbers come from.

• No longer use a multsig-type argument.
• Pull redeemScript from the wallet and determine the "m-of-n" for the multisig address
• With redeemScript information as well as "m-of-n" -> calculate the maximum inputs to keep the transaction around 15000 bytes or 30000 characters when signed due to Qt input box length limits and if the user argument of exceeds the calculated limit then ignore the argument.

I've made significant changes here and for the better imo. Since it makes no sense to use this feature to consolidate from a wallet that does not contain the redeemScript we just rely on the script being in the wallet. The script is added either when addmultisigaddress is done or addredeemscript is done. With new formula and calculations the transaction is within 100 bytes or less of actual end size which is an improvement for more accurate fees. This cannot be exact as serialization behaviour as well as signing can have different lengths on signatures due to many factors.

Yes, I have spelling errors to be corrected but for now putting this up for review while i get to those. Hopefully this will also address #1689

cheers

[div72]

Nice work, a couple of nits ahead:

[jamescowens]

See my comments. This is much improved over the original implementation I think.

[iFoggz]

should resolve everything.

[iFoggz]

will resolve the conflict once remote ssh setup to my box at my brothers place

[jamescowens]

ACK