WebDHT weparation of concerns (was Re: A Decentralized Hashtable for the Web)

On 11/03/2015 05:43 AM, Michael Bumann wrote:
> it sounds interesting. I am posting before I haven't fully 
> read/understood the spec, which maybe is not a good idea :)

Yes, please read the spec. :)

It helps us not repeat things that are already in the spec on the
mailing list. :)

> But somehow it reminds me of the Bitcoin or Namecoin block chain and 
> some projects that happen around there.

Yes, the goal is similar to Namecoin, except that this is not attempting
to take the Bitcoin blockchain and make it do something it was never
designed to do.

Part of our work here is to figure out the really interesting pieces of
the Bitcoin blockchain and reduce each thing to first principles,
implement those first principles as modules, and then construct new
things from those modules.

> Can it be reused?

Digital Bazaar's position is that we shouldn't re-use the Bitcoin
blockchain for something it was never intended to do.

> The Bitcoin keys are basically also some kind of identifiers. In fact
> the usecase of domain ownership is one that especially namecoin tries
> to solve - being a "decentralized open source information 
> registration [...] system". Easy speaking it solves the issue of 
> deciding about ownership of an unique name in a decentralized 
> system.

Yes, the downside being that you have all these other things that are
going on in the Bitcoin ecosystem that don't enable you to optimize for
that problem.

For example, the recent scalability discussions around block size in the
Bitcoin community have almost nothing to do with determining identifier
ownership. The WebDHT is not a blockchain by design, so it doesn't
suffer a number of the downsides that a blockchain design based approach
creates when dealing with identifiers.

To provide an example of how the problem could be decomposed:

The WebDHT has no memory. If a memory is desired, a separate ledger
format (aka blockchain) should be specified that can be written to /in
parallel to WebDHT operations that should be archived/. It could be
argued that this is the proper separation of concerns because you enable
the WebDHT (decentralized identifiers) to evolve at a different pace
than the ledger mechanism (long-term memory). There are certainly
downsides with that approach, but the point here is to have the proper
separation of concerns.

Don't get me wrong, I think Bitcoin is important and was a world
changing technology. That said, if we are to look at this from a systems
engineering standpoint, we have to understand the composable bits to
ensure that whatever we try to standardize will stand the rigorous W3C
standardization process. One of those hurdles we will have to overcome
is proving that we have the proper separation of concerns.

-- manu

-- 
Manu Sporny (skype: msporny, twitter: manusporny, G+: +Manu Sporny)
Founder/CEO - Digital Bazaar, Inc.
blog: Web Payments: The Architect, the Sage, and the Moral Voice
https://manu.sporny.org/2015/payments-collaboration/

Received on Wednesday, 4 November 2015 06:23:48 UTC