EML 2.2 units?

[twhiteaker]

According to EML Best Practices:

EDI has adopted the LTER Unit Registry and recommends that customUnit element be used for all units with content pulled from the Unit Registry, even when the unit is already listed in the standard unit dictionary.

So should all rows in EMLUnitDictionary have a custom value of false? Or maybe it's better if the R code uses for all units regardless of what the view says.

[gastil]

That decision is worth revisiting because Margaret overhauled the standard units in EML as part of EML2.2. In 2013 we chose to do only custom units partly because it was easier to code the EML generation to have just one way and not every unit is (was) offered as a standard unit. The Unit Dictionary itself was (and perhaps still is) not authoritative. Maybe EDI fixed this, but it used to be lots of bad-practice units were included.

[mobb]

that EML BP recommendation may be ... um ... misstated. IN MY DEFENSE, the doc did go out for review to 5 others. I hereby nominate @twhiteaker for the next go-round

[gastil]

Tim, just to be super clear, you do know the Unit Registry is a different thing than the Metabase table "EMLUnitDictionary", right? These are related, but different things.

[mobb]

agree that we might want to revisit (using simplified code). here is a link to the unit-registry work; those interested in where it should go next see the notes dir
https://github.com/EDIorg/unit-registry

[twhiteaker]

@gastil I was aware of this one:
http://unit.lternet.edu/unitregistry/
which I presume EDI was in the processing was cleaning up at https://github.com/EDIorg/unit-registry.

And that is different from the EML unit dictionary. I assume but have not checked that the unit dictionary in metabase comes from the EML unit dictionary.

[twhiteaker]

Speaking of best practice review, I did initiate a pull request for a markdown version of the BP back in August. I cleaned up some typos and XML along the way.

[atn38]

To revisit this, maybe we'd like to core-mb to ship with the EML 2.2 UnitDictionary instead of the one we have now? @mob, are EML 2.2 units in table form somewhere?