Bruce Momjian <bruce@momjian.us> writes:
> On Fri, Jan 10, 2014 at 08:15:53PM -0500, Tom Lane wrote:
>> Why would we change the operational procedure of "run this script
>> after pg_upgrade"? It just complicates life for users.
> If it is one command, why use a script? Just give them the command.
Because it won't necessarily always be just one command. I don't
think we need to complicate life for users (or even more, their
scripts) by making them actually read what pg_upgrade prints there.
The point of Peter's proposal is that the multi-step-database-analyze
process might be useful outside the context of pg_upgrade, so it's
worth extracting that as a utility. It doesn't follow that we need
to change the way users *of pg_upgrade itself* invoke the process.
regards, tom lane