> Attached v2 does as you suggest.
There is not a single test of "ctrl-c" which would have caught this trivial and irritating regression. ISTM that a TAP test is doable. Should one be added?
-- Fabien.