Paul Hammant's Blog: Choosing Between Blockchains And Vanilla Merkle Trees
This is a another new article in a mini-series that started with ‘Old-School’ Merkle Trees Rock, Merkle Trees In Pictures and Blockchains in pictures.
Say you were going to make an app or service, and colleagues are yelling “blockchain”. You should consider non-blockchain solutions too. You should also consider vanilla Merkle tree solutions. In my opinion, they are underrated and should be chosen more often. A table that might help:
| Criteria | Vanilla Merkle tree | Blockchain |
|---|---|---|
| Nature? | Tree-like adds, changes, and deletes (but moves are deletes and adds); everything is mutable | Chain-like notwithstanding an append-only restriction; nodes are immutable; with a merkle-tree as underlying an implementation aspect it is really part tree-like and part chain like |
| Node from which you can navigate to all other nodes? | The root. Address remains constant | The end of the chain. Location moves |
| Growth? | Suits a limited number of records, a percentage of which would change in a given period | Suits an ever increasing number of records, including acceleration of that over time |
| De/centralized? | Centralized - at least the root node is | Decentralized - consensus based authority featuring “proof of stake” and “proof of work” |
| History? | The subscriber can miss some changes to the tree, only “current” is purported to be available | Retained, notwithstanding forks, pruning, and lost transactions |
| Scope? | Most likely to be single purpose (say ‘2016 election results by voting district’) | Can be general purpose (e.g. Etherium). Specific purpose ones more likely though |
| Subscribable? | Is the appeal of this, including partial-tree subscriptions | Possible, logistically hard in the case of the biggest implementations, and ultimately unlikely by a non-government entity |
| Public/Private? | Either | Either, but ‘private-ish in public’ is regular too |
| Tamper evidence? | Yes | Yes |
| Tamper proof? | Tampering by centralized entity possible (or their delegates). | Only insofar as incorrect records could be appended, but that should be verifiable |
| Truthfulness of underlying payload data? | Tech does not speak to data’s accuracy, only to its authenticity | Tech does not speak to data’s accuracy, but via its “consensus” aspects makes a strong assertion about authenticity |
| Cost of initial implementation? | $ | $$ - $$$ |
| Worst case scenario? | Rate of change too high for subscribers to keep up | Dyson Sphere (Nat Pryce quote - see below) |
Nat Pryce’s Fermi Paradox / blockchain theory:
“Crypto-currencies are the reason for the Fermi Paradox. Alien civilisations discover crypto-currencies, start depending on them, require ever increasing amounts of energy to run their economy, eventually surround their star with a Dyson sphere to capture all the energy their economy needs, and effectively disappear from the universe.”
Thanks also to Prasanna Pendse for tweaks and fixes to the table above.