On Wed, 29 Nov 2017, Rob Myers wrote:
On 27/11/17 08:21 AM, Alan Sondheim wrote:
One question and query -
Who establishes the difficult algorithms miners solve? Are they
themselves generated within the blockchain? Is there a group that has
control over this?
The original Bitcoin difficulty adjustment algorithm was created by
Satoshi Nakamoto. Its is part of the C++ code of the Bitcoin client
software, so it exists outside the blockchain. Its stated objective is
to keep Bitcoin creating blocks of transactions roughly every ten
minutes. So the Bitcoin blockchain's ever-increasing energy consumption
is a result of the difficulty adjustment algorithm being a paperclipper
about targeting block time.
Changing the difficulty algorithm requires changing the Bitcoin software
source code. If that change is incompatible with the rest of the Bitcoin
network, clients running the modified software will form their own rival
network and start working on a separate "fork" of the blockchain. The
most notable fork of the Bitcoin software/network/blockchain at the
moment is "Bitcoin Cash", which explicitly includes a different
difficulty adjustment algorithm. Other chains also have forks over their
consensus rules, notably Ethereum and Ethereum Classic.
What I'm wondering, is who designs these algorithms? In other words, given
that this is an armored means of transaction, would there be a weak link
at the algorithmic end, or is the chain self-maintaining?
There have been experiments with putting the difficulty algorithm on the
blockchain. Eris did it at one point I think, and there was discussion
of this for Ethereum network but I don't think anything came of it
(although it does have "The Difficulty Bomb" to eventually enforce a
switch from Proof-of-Work to Proof-of-Stake). Decred, which I mentioned
in my last email to the list, makes voting on the parameters and code of
the difficulty algorithm part of the governance system built into the
software.
So control is usually with the software developers and embedded in the
code, although the developers lose the trust of the miners and of the
users another fork will pick them up, and there are experiments around
moving things on chain but that's not really happening yet.
Please explain this last paragraph - apologies for not understanding -
And a blockchain for cultural projects -
I wonder if it would be possible to establish an anti-currency, a
blockchain with easy mining solutions? For example, 2 50-digit numbers
multiplied? Does that make any sense? So that the currency built upon it
might stay worthless? But it could be used for cultural work, and would
still be stable?
Cryptocurrencies usually pay for the operation and securing of their
networks in their own currency, so making it worthless won't incentivise
people to do that. But there are systems like Hyperledger Fabric or
Tendermint that we could use to create such a network. Or if we accept
"almost worthless", we could use one of the faster/easier/cheaper
networks like Dogecoin (it's still going! :-D ).
This sounds fascinating, can you say more?
What kind of work would you use it for?
It's hard to say; I still am almost totally ignorant and would have to
work with someone who isn't. I'd be interested in thinking about the chain
in terms of codework, self-reflexivity, as well as partially embedding
text in it, that relies on the chain itself for at least part of the
content. I wrote about that in the codework article for American Book
Review, http://americanbookreview.org/issueContent.asp?id=37 - not sure
that's available now.
Thanks greatly,
Alan
- Rob.
_______________________________________________
NetBehaviour mailing list
[email protected]
https://lists.netbehaviour.org/mailman/listinfo/netbehaviour
New CD:- LIMIT:
http://www.publiceyesore.com/catalog.php?pg=3&pit=138
email archive http://sondheim.rupamsunyata.org/
web http://www.alansondheim.org / cell 718-813-3285
current text http://www.alansondheim.org/va.txt
_______________________________________________
NetBehaviour mailing list
[email protected]
https://lists.netbehaviour.org/mailman/listinfo/netbehaviour
