cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Marcus Crafter <>
Subject Re: AW: Some comments on Cocoon 2.0b2
Date Tue, 24 Jul 2001 12:40:20 GMT
Hi Carsten!

On Tue, 24 Jul 2001, Carsten Ziegeler wrote:

> > Stefano Mazzocchi wrote:
> >
> > ...
> > Yes, that's a solution, but I believe the design mistake here is that
> > "aggregation" is seen as "generation" while I see it as
> > "transformation".
> >
> > Suppose you have something like this:
> >
> >   generator -> layout page
> >   transformer -> aggregator
> >
> If I understand you right, you can already do this by simply using the
> generator you want and then using the CIncludeTransformer, or did I miss
> a point?

	This is amazing deja-vu, I was thinking about this just yesterday.

	Yesterday I was looking at the uses of aggregation, thinking it might
	be useful in our project. The idea behind it sounds great, but in
	practice I came up against some concerns.

	Take the simplest case - we have many HTML pages in our application
	which have 'headers' and 'footers', as most commercial applications
	probably do. Currently this is solved by included xsl templates, which
	are called in each html page before serialization.

	I thought aggregation would help here, removing the need to have these
	templates included and called, but after thinking about it for a while
	I unded up a little confused about whether aggregation could help.

	The keypoint I think is that aggregation solves the problem of
	aggregating multiple data sources together (which is extremely useful),
	but when one wants to aggregate multiple presentation components
	together it's difficult.

	Example, IMHO adding a header and footer to a document should be
	something done towards the end of a pipeline, just before serialization,
	not at the beginning, especially where long transforms are involved.

	We have some pipes where data is generated, created and then either
	formatted into html or csv. To aggregate the html part is difficult as
	we have to carry the header and footer markup all the way from
	generation through each transform, and then either use or discard it
	based on the final transform before serialization.

	It would be much better if we could aggregate in the 'header' and
	'footer' towards the end of the pipeline. This is kind-of possible
	using 2 pipelines, one to create the data, the second (invoked first)
	used to aggregate the header/footer together with the data from the
	first pipeline.

	But, as I said, perhaps I've not understood the full power of how to
	use aggregration.

	Any comments at all mate ? Is CIncludeTransformer useful for this
	purpose ?



     ,,$$$$$$$$$,      Marcus Crafter
    ;$'      '$$$$:    Computer Systems Engineer
    $:         $$$$:   Open Software Associates GmbH
     $       o_)$$$:   82-84 Mainzer Landstrasse
     ;$,    _/\ &&:'   60327 Frankfurt Germany
       '     /( &&&
           \_&&&&'     Email :
          &&&&.        Business Hours : +49 69 9757 200

To unsubscribe, e-mail:
For additional commands, email:

View raw message