I’ve just had an error event during DE CSV import, and I’m not sure if it is a bug or some kind of design:
I had a CSV file with 40 new data elements - what I did no know was that a few of them had shortnames longer than 50 characters.
I imported the file after setting “Reference mode” to “Deny invalid references”. This “Reference mode” is not described anywhere I’ve seen (not in the User manual at least), but from the name it seemed reasonable to assume that it was designed to drop e.g. any invalid catcombouid or optionset uid references - either by ignoring the whole row or by dropping the invalid reference value.
The import summary (attached) stated that 38 data elements had been created and 2 ignored. The “Conflict” part is a bit gobblygook as usual, but the reference to “shortname…” made it clear what the conflict was.
Based on this summary, I obviously expected 38 new data elements to have been created.
That was NOT the case - not a single new data element had been created, so all 40 failed (I checked first in the UI and then in the database itself). So the import summary code is buggy.
After shortening the offending shortnames, I re-imported the file using exactly the same setup, and the import summary showed 40 data elements created with 0 ignored and 0 conflicts.
This time the 40 new elements HAD been created.
Whereas there’s obviously some buggy code in relation to the import summary (reporting 38 created when 0 had been created), the question is what exactly this “Reference mode” is supposed to do. QUESTION:
Does “Deny invalid references” means that all imported objects will be ignored if one or more have an invalid reference, and does “Reference” in this case mean ANY field value in ANY row that causes a conflict?
Or is it supposed to only cause INVALID objects to be ignored?
I would like clarification on this, to determine if this is one or two bugs
46D Alma Road, 7700 Rosebank, SOUTH AFRICA
Tel/fax (home): +27-21-685-6472
Iridium SatPhone: +8816-315-19119