Roman,

> - Simple dynamic web sites. In most cases people need good
> - Embedded use. The direct competitor I guess is SQLLite, but on the
> - Small enterprises. We have almost everything we need. Users in
> - Medium enterprises. Now things are getting interesting. Hot
> - Large enterprises. Clustering, terabyte databases, high-volume

I think this is a pretty good market analysis. Combined with Geoff's
point about the oddity of IB being grouped with IDE's, it made me
think of the following:

It may be that FB's current sweet spot is the mid-market and that
this market is the most economically reached via distributors. Their
projects would be too difficult for a web-sale and too small for a
direct sales force with a 6-month sale cycle. Combined with Martijn's
point that IB seems profitable again, it may lead to a different view
of DevCo/IB.

Rather than the almost independent 35-person unit, it would be much
more intertwined with the sales and marketing effort for the IDE's,
both using the same two-tier distribution structure.

It may also explain why EnterpriseDB isn't more successful. I tend to
think of them as the combination of IBPhoenix, Janus, Upscene, with a
few added sales people and developers (based in India, I believe).
Oh, don't forget some venture money. Despite a year of coding up add-
ons, and another year of active selling, they have announced
remarkebly few success stories.

Maybe Borland have figured something out that has eluded us all this
time.

Paul