Against data generalizations

Ʌ: Thoughts

There’s actually no need to find over-arching hyper-standards that can capture every single use-case…

A common vocabulary is nice! (Hence my current (late 2020) obsession with RDF et al.) But the data need not necessarily be constrained to fit the format.

A data lake 🌐 approach is much more appropriate for most use-cases, where commonly understood data points are either marked, or “transformed out”, akin to “computed columns”, common in modern RDBMSs.