I read about prune option right now, actually i didn't hear about it before.
Yes this option can save some disk space but afaik first (awful N-days lasting) synchronization still requires to download full database.
My approach also cuts database and replaces all old blocks (except say last 6 blocks for security reason)
with series of blocks with rolled initial totals and optionally purged from tiny wallets crap (storing on six thousand current nodes and on the swarm of full wallets
 information that John have 100 satosi is too expensive for us and we may annually clear that balance as fee for miners or just delete).

So almost all nodes can hold only the rolled database (i can't estimate compression ration of the rolled database now, i am not advanced user as you can see).
 And only much less amount of archive nodes holds full expanded database.









 


2017-08-16 19:52 GMT+03:00 Nick ODell <nickodell@gmail.com>:
What makes this approach better than the prune option of Bitcoin?

On Wed, Aug 16, 2017 at 10:20 AM, Алексей Мутовкин via bitcoin-dev <bitcoin-dev@lists.linuxfoundation.org> wrote:

Let me describe the possible improvement of the bitcoin blockchain database (BBD)  size in general terms.

We can implement new routine : annual split of the BBD. Reason is that 140gb full wallet unconvinience.

BBD splits in two parts :
1) old blocks before the date of split and
2) new blocks, starting from first technical block with all rolled totals on the date of split.
    (also possible transfer of tiny totals due to their unprofitability to the miners, so we cut long tail of tiny holders)
3) old blocks packs into annual megablocks and stores in the side archive chain for some needs for FBI investigations or other goals.


Thanks for your attention,

Alexey Mutovkin













_______________________________________________
bitcoin-dev mailing list
bitcoin-dev@lists.linuxfoundation.org
https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev