(My feeling is it was working early well last year.)
Reducing the repeat_count at the moment creates inconsistent data in KoBoCollect. Except the repeat variable, nothing is changed, even count(), and even not on save or finalisation.
We couldn’t find any workaround so far, except to mark those cases manually (e.g. DELETE as person name) and clean-up later in re-edit / analysis, including all references.
And two further repeat problems, but with Enketo (pages style)
Hopefully, all three issues to be tracked in GitHub and solved soon.
Kind regards
