Postgres vs. MSSQL

Daniel C. dcrookston at gmail.com
Wed Nov 8 10:36:22 MST 2006


On 11/8/06, Jonathan Ellis <jonathan at carnageblender.com> wrote:
> That could mean anything from "Our DBA has looked into things and knows
> X and Y would be hellish to port" to "he's just scared of something
> new."

He's been trained extensively and formally on MS SQL, and informally
on Oracle.  Scared of new stuff does not describe him.

> There's essentially no reason to use MS SQL over PostgreSQL from a
> feature or performance standpoint.

If what you mean is that MS and Postgres are mostly equivalent in
performance and features (as opposed to "MS is not better", which
leaves "Postgres is better" as an option without explicitly stating
it) then I'm glad to hear it.

> - what version of MSSQL are you on?  if you're looking at paying for
> upgrades, moving to PG instead makes sense

No clue.

> - how invested are you in TSQL or other stuff that's a PITA to port?  if
> there's a lot, one option would be to say with MSSQL for your old apps
> and just do the new ones with PG.  (IMO this would make even more sense
> if you were also moving to django for new development.  No sense
> rewriting what ain't broke.  Assuming it is indeed in an unbroken
> condition ATM :)

This is probably what we'll do - write the new stuff in
Django/Postgres, keep the old stuff where it is.

> - how many developers and DBAs you'd have to retrain; if it's going to
> cost more in time lost than the new MSSQL licenses would, bite the
> bullet and open your wallet

Essentially none - it's just me (programming) and our lone DBA, who's
already familiar with Postgres.

As for the cost of switching languages - I learned PHP in about a
week.  I can't imagine it would take significantly longer to learn
Python, and the productivity gains from Django would be pretty
significant.  Are there good PHP frameworks that we could use instead?

Dan



More information about the PLUG mailing list