[
Lists Home |
Date Index |
Thread Index
]
[Jeff Lowery]
> > -----Original Message-----
> > From: Michael Brennan [mailto:Michael_Brennan@Allegis.com]
>
> > Although Ant is push-based whereas the pipeline proposal seems to be
> > pull-based.
>
> That's a good observation.
>
> I can see some flows where a mix of "me push" and "you pull" processes
might
> be hooked together: the main process pushing the primary inputs to
> subprocesses that are thereby triggered to pull secondary data to build
> their targets. I'm not sure how that would model in Ant or XML Pipeline;
> maybe as modular Pipeline scripts within a larger Ant framework? Must read
> more...
>
Anyone remember Morrison's "Flow Based Programming" (Van Nostrum Reinhold,
1994)? Fascinating stuff, which apparently he used very successfully at
IBM. Sort of a cross between recursively structured streams, pipes,
potential parallel processing, and pull processes. It ought be be well
suited to markup processing. I think there are still a few copies of the
book to be found (maybe used).
Cheers,
Tom P
|
