"cluster" scope
Alan Williamson (aw2.0 ltd)
Open source drives innovation and one area where the guys over at Railo
have been innovating is in the use of the "cluster" scope. We're now
being asked about this over in OpenBD.
In the absence of a CFML specification, I believe engine implementations
should become the defacto standard, and with that, I look to Sean and
Gert to lay down the foundations of what the "cluster" scope is and what
its characteristics are. After all, it is their baby.
I am not of course interested in implementation, as that is where we
engines compete. Instead i am looking at this from a perspective of the
CFML developer. What it means to them.
In the spirit of openness and evolving the CFML langauge, I am not
interested in merely cloning the "cluster" scope, but buying into
Railo's vision and mindset for it so we too can offer the same feature
and not fork off another feature.
Thanks
a
Sean Corfield
<al...@aw20.co.uk> wrote:
> Open source drives innovation and one area where the guys over at Railo have
> been innovating is in the use of the "cluster" scope. We're now being
> asked about this over in OpenBD.
Thanx.
> In the absence of a CFML specification, I believe engine implementations
> should become the defacto standard, and with that, I look to Sean and Gert
> to lay down the foundations of what the "cluster" scope is and what its
> characteristics are. After all, it is their baby.
We're discussing this right now because we're leaning more toward the
ability to overlay shared scopes on named caches as the better
solution to the clustering problem, e.g., aligning server or
application scope with a named, distributed cache. Custer scope
pre-dates my time on the Railo team so I'm not best placed to comment
on its vision beyond that.
--
Sean A Corfield -- (904) 302-SEAN
Railo Technologies, Inc. -- http://getrailo.com/
An Architect's View -- http://corfield.org/
"If you're not annoying somebody, you're not really alive."
-- Margaret Atwood