On Tue, Jan 30, 2018 at 12:48 PM, David G. Johnston < david.g.johns...@gmail.com> wrote:
> On Tue, Jan 30, 2018 at 8:34 AM, bto...@computer.org < > bto...@broadstripe.net> wrote: > >> >> When this procedure got old, I started using a script created using >> pg_dump and pg_restore, as initially outlined here: >> >> > Yeah, the short answer is PostgreSQL doesn't make it possible to edit > "middle" views without having the code on hand for all dependent views so > you can recreate them. You either maintain those views and order manually > or you rely on pg_dump to figure it out for you (the former, with version > control, is highly recommended). > > I could see it being possible to program the server to be more helpful > here - by say allowing it to drop but remember view definitions and the > re-create them from the remembered versions by name - but no one has seen > the motivation to do so; I suspect partially in light of the fact that > "version control" is a recommended practice. > > David J. > > *Just a side note, it is not a good practice to create views based on other views. * *Multiple reasons are stated here:* *https://dba.stackexchange.com/questions/5487/is-nested-view-a-good-database-design <https://dba.stackexchange.com/questions/5487/is-nested-view-a-good-database-design>* *To summarize, you suffer performance degradation, columns names and the tables referenced become obscured and you incur the problems you are now experiencing.* -- *Melvin Davidson* I reserve the right to fantasize. Whether or not you wish to share my fantasy is entirely up to you.
