forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Steven Noels" <>
Subject RE: Graph data
Date Thu, 21 Feb 2002 11:37:12 GMT
John Morrison wrote:

> > Even though the (Caching)CInclude version of the pipeline
> > does an extra XSLT
> > transformation (there's no XPointer support in the
> > CIncludeTransformer), some
> > rough tests with JMeter told me that the CInclude version is
> > twice as fast as
> > the XIncludeTransformer + I see my cache being filled.
> I wonder why it's twice as fast...

Using the XIncludeTransformer+XPointer, each daily log file is subject to
parsing and XPath processing (to include unwrapped datum elements from the daily

Using CInclude+XSLT, the daily files are included as-is, no unwrapping of datum
elements: this is done in a separate XSLT transformation on that single big
concatenated document.

Plus there's cacheing, less filesystem-calls etc...

> > Would you folks (John?) like me to patch my submission of
> > yesterday after it
> > gets checked in somehow, or just send in another zip with
> > different filenames
> > etc...?
> I'd *like* to setup a proper file system structure in cvs for
> the webapp - unfortunately I have no idea as to what is
> required...

Well, we could at least start to set up an infrastructure for our own project
pages, built using a CLI-Cocoon: ?

I'd say we start off with the static stuff described in Stefano's [RT]: primarly
project pages, stylesheets, and the graph logs...

Can we start with the assumption there will be one instance of a forrest/cocoon
webapp, presumably triggered through the CLI, that collects a number of
documents from the different project dirs (forrest included), and pipes these
through some stylesheets generating the static site Stefano sketched in
xml-forrest\layout\\ ?

Ah, I just see David addressing our shared concern. Who will take the lead in

Fingers itching to get started :-)


View raw message