Firstly, thanks to Jakub for the translation.

Secondly, the results are obviously not real world results. If so,
absolutely no-one would use MySQL, and PostgreSQL would be even less
popular. I am not suggesting the author had any bad intentions,
perhaps he just chose by coincidence queries that Firebird handles
well and the others dont. I know that we would dismiss the paper if it
compared a select count(*) query.

Perhaps the PHP-Firebird components are more mature than the ones he
used to connect to the others (although I doubt this could be true for
MySQL, LAMP). Maybe the interface layer caused a bottleneck.

More likely though, it is a testament to the self-tuning-ness of
Firebird. Straight out of the box, Firebird generally has good
performance. While there are a few parameters that can help in certain
configurations, this is one of its key strengths. The hardware is
relatively modest (in a DBMS server sense), and that is another
strength of Firebird.

I do love it when Firebird comes out on top of benchmarks, but not at
the expense of the integrity of the conclusions reached. Perhaps a
larger test using a lot more sample queries, complex joins etc
repeated across several different interfaces PHP / C / Java / .Net /
Delphi etc would give a more accurate picture of dbms performance.

Adam

--- In Firebird-general@yahoogroups.com, Lester Caine <lester@l...> >
I use PHP heavily now in distributed apps, I *AM* interested in how the

> time differences are achieved. Especially if it will provide more
> information for my papers in two weeks time ;)
>
> While I don't really have the time, perhaps I should install a copy of
> MySQL and Posgres and run the bitweaver site on each to compare the

results!

>
> --
> Lester Caine
> -----------------------------
> L.S.Caine Electronic Services
> Treasurer - Firebird Foundation Inc.
>