I had been thinking that this was a unexpected side effect of r14297, but local testing generates the same query, so that theory is out.
This means its one of our ongoing 'data changed slightly, query becomes intransigent' problems.
I had been thinking that this was a unexpected side effect of r14297, but local testing generates the same query, so that theory is out.
This means its one of our ongoing 'data changed slightly, query becomes intransigent' problems.