Re: Reporting script runtimes in pg_regress

John Naylor <john.naylor@2ndquadrant.com> writes:
> On 2/15/19, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:
>> We should also strive to align "FAILED" properly.  This is currently
>> quite unreadable:
>> 
>> int4                         ... ok (128 ms)
>> int8                         ... FAILED (153 ms)
>> oid                          ... ok (163 ms)
>> float4                       ... ok (231 ms)

> If I may play devil's advocate, who cares how long it takes a test to
> fail? If it's not difficult, leaving the time out for failures would
> make them stand out more.

Actually, I'd supposed that that might be useful info, sometimes.
For example it might help you guess whether a timeout had elapsed.

            regards, tom lane