At 02:11 AM 8/3/00 +0100, you wrote:

> > From: Paul Gallagher
>
>Hi,
>
> > I agree. Its very important not to look at speed only.
>
>Although I agree, It is the sort of thing "managers" look for when choosing
>products, technical people look at features.
>
>Why I posted this link was there was a IB - PostgeSQL head to head a while
>ago on a web site (can't remember which one), and IB came a poor second.
>This new one shows it is at least twice as fast on their tests.

If I remember correctly that was a ZDNET (PCweek?) test. What it showed was
that under a single user load IB was up to three times faster then
postgres. When the user load went up postgres pulled ahead because it was
using a shared cache and IB was not. I think this test was using a older
version of IB and postgres. I am not sure if that shared cache has been
implemented with superserver or not anybody know for sure? My reading of
the benchmark was that the core engine is faster then postgres but there
seems to be some room for improvement in memory management.

Since then both IB and postgres have been majorly overhauled. It would be
interested in running those tests again. Does anybody know of an open and
fairly unbiased benchmark I can play with? Seems like every benchmark I
have seen seeks to prevent you from telling anybody else about the results.
Sure databases are tricky to tune and apps are tough to write but still we
ought to be able use some sort a comparison if for no other reason then to
figure out what could be improved.
----------------------------------------------
Tim Uckun
Mobile Intelligence Unit.
----------------------------------------------
"There are some who call me TIM?"
----------------------------------------------