On Fri, Mar 23, 2018 at 11:39 AM, Ben Finney <ben+pyt...@benfinney.id.au> wrote: > Chris Angelico <ros...@gmail.com> writes: > >> There is NOT always a good reason for a suboptimal configuration. > > True. Did anyone claim otherwise? > > What I saw Steven responding to was your claim that there is *never* a > good reason to do it. > > To refute that, it's sufficient to show that good reason can exist in > some cases. That's entirely compatible with a good reason not existing > in other cases. >
I'll concede that sometimes it's a lot of effort to fix misconfigured databases. However, it's a form of technical debt, and when you start investigating problems ("this isn't working because X is incompatible with Y"), that's an indication that it's starting to cost you to *maintain* the misconfiguration. Hence it's more likely to be worth just paying off your technical debt and moving forward. ChrisA -- https://mail.python.org/mailman/listinfo/python-list