In 2015 everyone started getting excited about “the blockchain”. Part of this interest is simply amateur hour, as technically ignorant people discovered “write once memory” without understanding what we already know about the uses and limitations of this concept.
Hint: we have had log file systems for decades, they are a little used option in most database management systems. They are useful for some things, but they are pretty clumsy and inefficient for most purposes, which is why we use the other storage methods offered in DBMSs.
In any case, “write once” storage can’t possibly solve the “garbage in, garbage out” problem: you need to have good data and a chain of trust end-to-end. Storage management isn’t really the key challenge.
Nevertheless, interest in the blockchain has also been driven by the dawning realization that the transparent shared ledger component of cryptocurrencies is very useful, even without the currency and mining component. People are taking the cover off of ‘Bitcoin’ as described in the Nakamoto document, dismantling it, and remixing the pieces in different ways. This process may be controversial for fundamentalist Nakamoto-ites, but it is the sign of a vibrant technology.
The biggest new thing in 2015 has been the explosion of interest in “private blockchains” and other variants, which unbundle the write-once, distributed data structure from the ‘scratch off lottery’ mechanisms of cryptocurrency mining. Private blockchains also decouple blockchain applications from the Bitcoin cryptocurrency itself. There is a happy side-effect, because private blockchains may stand aside the ideological baggage of hard-money rhetoric and intimate associations with black marketeering.
Preson J. Byrne comments at Coindesk on “4 Hype-Free Predictions for Private Blockchains in 2016”. His main points are:
- Nobody will own the stack
- Blockchains will be accepted as general-purpose databases
- Mining will be (mostly) relegated to irrelevance
- Code, don’t talk
Well, number 1 and 4 are self evident. As he says, anyone who even asks about a winner takes all domination of blockchain technology “betrays a degree of ignorance about what blockchains actually do.” And we’ve had enough exciting talk, let’s implement something. (Patent filings shows that this is underway.)
Byrne’s second point is not self evident. I would note that most full service database management systems have offered write-once files (AKA log file systems) for a long time. Most applications don’t use them, because they are awkward and slow. Blockchains are even more awkward and slow, and will have limited use. We’ll see what they are actually useful for.
Point 3 is quite interesting, and absolutely dead on in my opinion. Nakamoto style mining is needed only if there are no trusted nodes, and the network is completely flat—end users connecting to each other in one hop. Most distributed systems are hierarchical, and require trust relations. In these cases, you don’t mining, and good riddance! I, for one, am tired of hearing about Bitcoin prices (i.e. exchange rates) and the technical needs of digital traders.
Of course, things are not without pitfalls and troubles. However magical blockchain may be, it can’t overcome the realities of corporate competition. While IBM and some collaborators donated it’s technology (and support) to the Linux Foundation (notably not the Bitcoin Foundation), Blythe Maters faced pushback from fellow Wall Streeters, apparently unwilling to have their pockets picked by JPMorgan. (Wall Street is still Wall Street.)
At the same time there have been numerous cunning plans to use blockchain certificates for many purposes, including recording educational certificates, provenance of digital art, and various financial instruments. These efforts remain to prove out, if only because they are solutions to non-problems.
Application of the technology to non-digital assets is proving challenging. Provenance.org is working hard to use blockchain technology to document supply chains. It’s taking a while to get booted up, which is not at all surprising. I have a certain amount of optimism about this project, if only because they clearly understand and are tackling the end-to-end nature of the problems.
Another effort has been a much hyped plan to build a land title registry for all of Honduras, using the Bitcoin blockchain. It’s not clear to me that this is the most critical problem Honduras needs to solve, but it’s an interesting idea. Unfortunately, the problem appears to be stalled, most likely due to real world politics. (Is it possible that the current unsatisfactory system favor the interests of the wealthy and powerful? That the problem is political, not technical?)
In short, however revolutionary blockchain technology may be (and I think that is debatable), getting it out into the real world is going to be difficult. The powers that be do not want to be “disrupted”, and they do not have to sit idle.
But now that blockchains have been liberated from the tyranny of Bitcoin, progress can be made in 2016.
Cryptocurrency Thursday
Robert McGrath's Blog 