Re: MySQL+InnoDB vs. PostgreSQL test? - Mailing list pgsql-advocacy

In an attempt to throw the authorities off his trail, JanWieck@Yahoo.com (Jan Wieck) transmitted:
> Josh Berkus wrote:
>> I've had requests from a couple of businesses to see results of
>> infomal MySQL
>> +InnoDB vs. PostgreSQL tests.    I know that we don't have the setup
>> to do full formal benchmarking, but surely someone in our community
>> has gone head-to-head on your own application?
>
> how does someone compare an Apache+PHP+MySQL "thing" against something
> implemented with half the stuff done in stored procedures and the
> entire business model guarded by referential integrity, custom
> triggers and whatnot?
>
> Seriously, I am tired of this kind of question. You gotta get bold
> enough to stand up in a "meeting" like that, say "guy's, you can ask
> me how this compares to Oracle ... but if you're seriously asking me
> how this compares to MySQL, call me again when you've done your
> homework".

Actually, before saying anything in public about their products, check
out what they require for use of their protected trademarks.
<http://www.mysql.com/company/trademark.html>

To wit, they indicate that:

  "The MySQL AB Marks may not be used in a manner or with respect to
  products that will decrease the value of the MySQL AB Marks or
  otherwise impair or damage MySQL AB's brand integrity, reputation or
  goodwill"

It seems to me that presenting a benchmark that did not favor their
product could be quite reasonably considered to be an "impairment" of
their integrity, reputation, or goodwill, and therefore be something
worthy of legal attack.

That certainly wouldn't be new to the database industry; numerous
(most?) database vendors forbid third parties from presenting
benchmarks without their express consent.

It is actually rather surprising that despite having the budget to put
together a "benchmarketing" group, the only results that they are
publishing are multiple years old, with a paucity of InnoDB(tm)
results, and where they only seem to compare it with ancient versions
of "competitors."

And, of course, if "MaxDB(tm)" is their future, replacing the older
storage schemes, the benchmarks should be based on that.  And the
benchmarks that exist there are all based on the R/3 (tm) SD module,
which is spectacularly different from the usual web server work.  (It
looks like that involves throwing a load of BDC sessions at the
server, but I'm guessing, and in any case, it's work SAP AG did...)
--
let name="cbbrowne" and tld="acm.org" in name ^ "@" ^ tld;;
http://www.ntlug.org/~cbbrowne/nonrdbms.html
"Has anyone ever thought about the fact that in general, the only web
sites that are consistently making money are the ones dealing in
pornography?  This brings new meaning to the term, "obscene
profits". :)"  -- Paul Robinson <postmaster@paul.washington.dc.us>