Subject Re: [Firebird-Architect] Re: Database Culture and Progress
Author Jim Starkey
David Johnson wrote:

>I think I understand your frustration Jim, but I think it may be
>I have watched this forum for a few months now, and I am just beginning
>to get a grasp of your long-term architectural vision. You have
>introduced bits of the architecture one at a time, and they have been
>debated piecemeal until a resolution of sorts was reached.
>I suggest that, without a full disclosure of your aims, you will
>encocunter more roadblocks than assistance. You are getting into
>features that are interdependent, and any part of the interdependent
>features without the a grasp of the intent of the whole will not make
>sense, hence the resistance.
Thanks for your thoughts. Yes, I have a clear ideas of where I think
Firebird ought to go. Many are at the implementation level; I doubt
many folks still believe I've been reticent about what I think should be
done. I thought I had been clear about where I thought new
functionality (a very DEC word or non-word) should go, but I doubt I
have put it down in black and white.

Ann and I are about to disappear to Maine on our sailboat for three
weeks, so this isn't the best time to strike up a full discussion of
"the vision thing". For now, suffice it to say that I believe that the
future belongs to three tier-ed applications (aka application servers)
with the browser as the universal platform. I've more than a fair bit
of experience in this space with Netfrastructure. Netfrastructure and
Firebird are not competitors -- Netfrastructure does thing have Firebird
can't (for 17 years, at least), and Netfrastructure will never open
source. That said, I believe Firebird can and and should be the
database of choice for application servers that need an open source
solution. To succeed, Firebird needs to address a number of serious
performance related problems, though this work is well under way. But
it also has to address issues specific to the Web, the most significant
computing platform in history. The Web is flexible, dynamic, and must
support both search centric and drill down interfaces simultaneously.
Contemporary database systems are marginally adequate for drill down but
next to useless for search.

If you want to get a clear of idea of where I think we should go, walk
around the Netfrastructure site. The template engine is off limits, but
the rest of the technology can be borrowed or stolen. The
Netfrastructure database model can't be retrofitted to Firebird for
backward compatibility reasons, but many other features can. One
critical feature is filtersets, which I intentionally chose not to
protect with patents. Filtersets, which do require significant
infrastructure to support, are probably the most important database
concept in this decade.

The current Firebird users are locked into client/server computing and
have chosen Firebird primarily because the price is right. The future
users need us to push the technology envelop to address their needs.
They aren't here now because we can't meet their needs, but they're our
hope for the future.

>Have you considered using LDAP or an LDAP like structure for
I've got a customer with an LDAP server layered on Netfrastructure, but
personally, I haven't gone much beyond learning to type LDAP. Maybe if
I have a chance to catch my breath...