We are having a problem with using subqueries on persistent objects, to which a parametrized filter is applied. We are observing a phenomena where values from the filter and values from the query are switched, so that I have a filter value A, and parameter value B, but in the actual SQL (observed w/ P6Spy) the value or the filter passed is B, and the query parameter is A.

In our scenario where the main query and the subquery are executed against the same table (and therefore the same persistent object type), which uses the filter (and this is the desired behavior), but the parameter and filter values are switched (binded in random order as specified above), in the resulting SQL. This is clearly a bug and should be fixed. Could you provide some timeline for when this will be addressed?

I've read JIRA issues HHH-512, and HHH-530 (and commeted on those), but this issue seems to be a bug. The query worked in 2.1, and stopped working when we added a filter to this object.

Also, I've experimeted with changing order of the parameters in where clause, and observed that parameters were binded in different order as well...