The beauty of an open source, BSD-licensed project like PostgreSQL is
the entire "Who cares?" possibility list.
If you have a Windows shop and you have Windows trained personnel, then
you can use PostgreSQL.
If you have a Linux shop and Linux trained personnel, then you can use
PostgreSQL.
If you have a FreeBSD shop and FreeBSD trained personnel, then you can
use PostgreSQL.
I think a picture is starting to form here.
Monkey-wrench time...
Suppose that I have a 4-way AMD64 Windows system running PostgreSQL and
even that runs out of steam. I have added as much ram as the system
will hold and the load is still causing problems.
Now, I can get an IBM machine running SUSE with a pile of processors and
gobs of ram and scale to whatever TPS I need.
And the data + schema?
Dump from the Windows box, load on the IBM SUSE box.
I might even be able to SLONY it over without ever going off line.
IOW -- what't the whole point of open source BSD licensed projects?
It's that you just do whatever you like to solve the problem in the way
that is best for your organization (with your personnel and your
hardware and your training and your data). And if you need to scale to
somewhere else, then you can do it.
It's the best of all worlds.
