plinehan wrote:

>
> [IB] wasn't generating
> the return on input that was being demanded
> ... nobody could figure out a way
> to make IB a big enough loser of money so
> that it was perceived that it was riding the
> right wave - in surf-dude speak?

An Irish surf-dude? Whoa... scary. But yeah,
that's about where things were.

>
> You're basically saying that Borland was trying
> to run with the "dot-com, internet-thang", but
> didn't know which ball to choose, so chose
> IB because it might (while still profitable
> going forward) not have been the brighest
> star in their galaxy?

Err, it was a little messier than that. The
original idea was to "sunset" the product - stop
development, cancel V6, and milk support contracts
until everybody noticed that that the cow was dead.
That's when Paul Beach (then GM) and several other
managers quit. Their departure got the attention
of several InterBase partisans and the planned quiet
execution turned into a noisy mess. The foray into
the open source / dot com stock bubble was making
lemonade out of the lemon they'd laid.

>
> I would have hardly considered IB as
> a secondary asset.

They did.

> Do you know (and are
> you in a position to say?) what sort of
> money was being spent and made by the
> Interbase division of Borland/Inprise
> at the time?
>

Unh, no.

>
> ...trying to
> explain why (despite its superior locking model)
> [IB] never made it to the big time.

The database market is pretty tough. Oracle
has always maintained very good account control,
and Microsoft is pretty good in that area too.
Borland may very well have had some unpleasant
discussions with them on the subject of pushing
InterBase and the impact that would have on their
access to early releases, free copies, and all
the other tools you need to develop data access
tools.

>
> I would love if you could let me know if there
> are any howlers in there!

I think I mentioned a bit of history - but largely,
your account is fair. InterBase started in the same
month as Sybase, but the Sybase people had done one
startup before - we were complete novices. They got
a lot of venture and had full-page three-color ads
before they had working code. As it turns out, that
was the winning strategy - somehow magazine seem to
notice companies that buy advertising space and write
glowing articles about those same products. No
advertising budget -> no media coverage.

Not that we didn't screw up in a lot of other ways...

> Basically, I say that
> Borland were saying to clients something like
> "TurboPascal/Delphi can work with any db, and,
> oh, yeah, we have one that you might consider...".
>

See above for a reason they might have taken that
attitude.

>
> As the Arch-Doyenne of IB/FB, I would be grateful
> if you could contribute your opinion to the
> debate between Paul Ruizendaal and myself over
> what is a good/reasonable metric for the analysis
> of db penetration in a given marketplace.
>

Err, that's another topic on which I have absolutely
no insight, except to say that there's less information
available than you may think. You can't find Oracle's
database sales from their financials. Their pricing
model is "everything you can possibly spend plus 20%",
so the cost per copy depends on the depth of the
pockets of the customer. Hey, I'm not complaining,
that's the model we used too, though less skillfully.

As for using help wanted ads, well, Oracle is a very
specialized skill. Anybody who can code and knows SQL
can use Firebird effectively. Eats the hell out of the
tutorial and certification market, but still, easy to
use is a good thing (tm).

My feeling from the business we're doing at IBPhoenix,
the activity of the lists, and the mentions of Firebird
in articles about open source is that we're growing fast.
Nothing succeeds like success. Noticing and promoting
our successes is also a good thing (tm).

Cheers,


Ann