3 ms·
If you accept that for many uses of publicly verifiable data structures that centralization is actually OK (and actually often a good thing, as it matches well
by aeijdenberg 10y ago
If you accept that for many uses of publicly verifiable data structures that centralization is actually OK (and actually often a good thing, as it matches well how most democratic societies function), then there are other simpler alternatives to proof-of-work that still give many of the same benefits.
See for examples the append-only logs used in Certificate Transparency (https://www.certificate-transparency.org https://www.certificate-transparency.org), and a more generalized hosted form provided by Continusec (https://www.continusec.com https://www.continusec.com) that builds on the same principles. (disclaimer: founder of Continusec)
- runeks 10y agoThe difference is that this data structure defines who's rich and who isn't (literally). There is a whole lot more incentive to tamper with such a data structure than basically any other data structure in existence.