Has anyone else seen cases where errors from subselects are not being
raised? I just had it happen again but can't find a simple case that
reproduces it. The query is like this:
insert into ...
select (a bunch of CASE statements specifying a subselect to use if the
value of the attribute is null)...
from...(a messy JOIN)
in which one of the CASE contains a subselect which has an "Attribute
not found" error when executed independently. However, the only way it
manifests itself is through a NOT NULL constraint not being satisfied in
the insert.
Should I spend any time trying to reproduce this in a simple case, or is
there some known stupid way of making this happen?
Thanks!
