Closed xrotwang closed 3 years ago
Thinking a bit more, I'm retracting the proposal. One design principle of CLDF was to only add things that have clear use cases in analysis - rather than aiming at some sort of "completeness". Judging from the clld experience, "units" will probably end up as a catch-all category, making automated re-use almost impossible.
The
clld
data model distinguishes two kinds of "measurements" or "values":The latter was introduced to cover cases like constructions, or paradigms. But it could also be used to model things like allophones (see https://github.com/cldf/cldf/issues/93#issuecomment-765235046). Now we already have components modeling particular units of language, namely
FormTable
andExampleTable
. So the question is: Would a genericUnitTable
make sense - or should we add more specific components, modeling specific units, when the need arises?