dhis2-devs team mailing list archive

[Lars Helge Øverland <larshelge@xxxxxxxxx>]

>
>
>
> Although again a little off-topic to the original post, the reason why
> this happened is still not 100% clear to me, but seems to be have been
> introduced with a script which I wrote to convert datasets to data element
> groups (so that they could be used in MyDatamart). I have previously posted
> some R code which will generate DHIS2 style UIDs, and I suspect that this
> code introduced the duplicate UIDs, which proved to be OK in 2.10 and 2.11,
> but not in 2.12. Of course, if I had checked this before creating the data
> element groups (which I didn't) this would have never happened. But I find
> it strange that there was not a unique constraint on this key in 2.10.
> Again, maybe a result of  hack perhaps on my part, or just an oversight in
> the schema. So another lesson if you are gonna hack DHIS2, be careful about
> relying too heavily on the constraints in the schema!
>
>
Yes it was no doubt a mistake not to include uniqueness constraints on uids
from the start, would have saved some trouble. It was not a conscious
decision, rather an oversight. At least we have it now.

Lars