Andrew Dunstan <andrew.dunstan@2ndquadrant.com> writes:
> On 06/19/2018 05:21 PM, Andres Freund wrote:
>> Given that we've "officially" stopped relying on context diffs,
>> I don't see why we'd not also retire them in pg_regress.
> Well I suspect at least one person would be made unhappy ;-)
Not requiring them is considerably different from actually trying
to stamp them out. To my mind, where we're trying to go here is
not having one set of people impose their diff format preferences
on other people.
There's already a (perhaps underdocumented) way to make regression
diffs look the way you want in local runs. Andrew's idea of getting
the buildfarm server to display diffs either way would go a long
way towards having that same flexibility for buildfarm results, though
it does seem like rather a lot of work for a small point :-(
regards, tom lane