On 23 Nov 2001, at 19:02, lester@... wrote:

> > The real idea here, is that there are complex methods of
> > maintaining an auditable series, and there are simple methods, so if
> > you know a simple one will work with what your doing, why use a
> > complex method.
>
> Agreed - But
>
> I think the point people are trying to make is that Firebird
> is ****** stable. Anything that gets added to it must not
> break that stability. We can make it do even more but for
> some people it is fine as it is. So anything that is added
> must not break existing applications and people can use new
> features if they want to.
>
> Returning to the Subject in question - we are already
> diverging from IB, so whether IB7 incorporates FIRST amongst
> other things is another matter and another discussion. The
> Idea for the new field has merit provided it does not break
> such things as IBO's generator 'manager'.
>
> In the process of the discussion we have added 'sequential'
> generators - to isolate that discussion from the simple
> AUTOINC, and RETURNING. Three wish list items, which
> hopefully will appear in the Firebird wish list.
>
> Have I missed anything?
>
> If not, then now is the time to start thinking about other
> wish list items <g>
>

In Unix they call this a fork, where one project turns into two,
projects that then go their separate ways, it's whether the PHB's at
Borland let the marketing guys to convince the developers that
trying to keep close enough to FB to stay compatible is a good
idea. I say this because FB is now in the lead, and a diff on the
features list shows what users and developers want rather then
what a bunch of executives think they want.

The ideal now is to discover a way to add new features, without
losing the stability. However we also need to not be afraid to dump
bad features when we can come up with better ones. The services
API is a good example of something that needs to be replaced.

Paul







Paul Schmidt
Tricat Technologies
paul@...
www.tricattechnologies.com