On Wed, Apr 6, 2016 at 3:55 AM, Alexey Bashtanov <bashtanov@imap.cc> wrote:
> Hi all,
>
> I am searching for a proper database schema version management system.
>
> My criteria are the following:
> 0) Open-source, supports postgresql
> 1) Uses psql to execute changesets (to have no problems with COPY,
> transaction management or sophisticated DDL commands, and to benefit from
> scripting)
> 2) Support repeatable migrations (SQL files that get applied every time they
> are changed, it is useful for functions or views tracking).
If you are using COPY, then you must be loading data (not just
changing schemas).
But if you are loading data, how can you have it be repeatable? How
would it know which data needs to be (or not be) loaded again? Do you
want it do a three way comparison (The current patch it is trying to
apply, the most recent patch that had been applied, and the live
database) and resolve conflicts?
Cheers,
Jeff
