Instead of carrying full blockchain, why not this solution?

[bitpump]

Instead of carrying full blockchain, why not this solution?

Summary:

client will include only last 1.000 full blocks and 10% of all other blocks.

Details:

for last 1.000 blocks : 100% include

for all other blocks (before these last 1.000 blocks) decision either
 A. to include or
 B. storing the following info from a block
        - block number
        - hash of block
        - list of (first 16 characters of) all bitcoin addresses impacted by this block

choosing between A and B is based on some randomness and if it has an acceptable distribution (10%?) on the network (using a Distributed Hash Table, just like with bittorrent?)

now, if a client wants to know the "balance" of a specific bitcoin address, it can just check which blocks to download (P2P) to provide a solid confirmation of this "balance"

to get an immediate non-solid indication of a "balance" of an address, consumers can use a trusted party website (which has downloaded the full blockchain)

[R2D221]

Isn't this what pruning already achieves?

[R2D221]

to get an immediate non-solid indication of a "balance" of an address, consumers can use a trusted party website (which has downloaded the full blockchain)

Why not SPV?

[bitpump]

Isn't this what pruning already achieves?

thanks for pointer!
http://bitcoin.stackexchange.com/questions/10333/blockchain-long-run-issue/10334#10334

so is pruning already being used / implemented?
if so, why do I keep seeing this argument (size of blockchain) being used?

[bitpump]

to get an immediate non-solid indication of a "balance" of an address, consumers can use a trusted party website (which has downloaded the full blockchain)

Why not SPV?

thanks again for the education,
so apparantly, a "thin" and a "thick" client already have been designed in the original bitcoin whitepaper
https://en.bitcoin.it/wiki/Thin_Client_Security#Simplified_Payment_Verification_.28SPV.29_Clients

I should do some more homework before asking more questions like this
Thanks for your time !

[Nrcewker]

because the outcome plan involves the redistribution of profits, vested interests will not agree

[NorrisK]

because the outcome plan involves the redistribution of profits, vested interests will not agree

Why would profits be redistributed? It only shares the storage space for archived blocks. Still 10% of all nodes will publish any specfifc block and all would publish the last 1000 blocks. Doesnt look like a bad idea imo.