Antonio Soler has a number of sonatas for which “Keyboard” has yet to be added to the Instrumentation data item. See below:
And yet, it is clear from the canonical name that these sonatas are for ‘keyboard’. There is no mystery here except to wonder why the instrumentation column isn’t complete.
Scarlatti has another slew of Keyboard sonatas with missing Instrumentation values.
So, I ask again:
- Do you have any intention of finishing this composition DB project so that its values can be actually used?
- If you have gone as far as you wish, will you open the DB to user edits, at least as far as his/her own library is concerned?
- If you won’t finish or allow us to, will you announce that this is a dead project so we can stop being teased?
There are other compositions I could list for you that have no date or period or form. While some of these are legitimately empty, most of the empty just aren’t … finished.
Thanks for the report–this is a display bug. These compositions have nothing in the instrumentation field, but a display rule is trying to display " for " anyways. Simple to fix, will be fixed in the next release. They should just say “Sonata”.
I am not sure what the “composition DB project” you’re referring to is or what it would mean to finish it. Can you explain?
Not “Sonata for Keyboard” or “Suite for Orchestra” or more generally, for ? That would be sensible.
I am talking about Roon’s composition DB, which lists Composer, Date, Period, Form, Instrumentation, and Performances. There seem to be numerous missing classifications and inconsistencies. But there may be good reasons for the absences.
For long-completed classical works, is this DB finished or is there still work to do?
We don’t do any hand-grooming or generation of metadata here. There is no Roon Composition DB–all of that data either comes from upstream sources or automatic processes. We have no infrastructure for editing this data ourselves, and there is no project to “complete” it.
That “X for Y” treatment corresponds to " for ", so it’s a display of two independent fields, each of which may or may not be present in the underlying data, so the behavior is going to depend on whether they have values populated or not.
There was another bug here preventing the instrumentation field from being populated properly for TIDAL/Qobuz works–this will be fixed.