On Wed, 30 May 2001, Donald Ball wrote:

> On Tue, 29 May 2001, Berin Loritsch wrote:
>
> > If it doesn't resolve to to the root of the servlet context, then the
> > protocol shouldn't be called "context".  By making things sitemap
> > relative, you are diluting the concepts of the Servlet spec.  You are
> > also increasing (IMO unnecessarily) the learning time required to
> > understand Cocoon.
>
> _thank you_. i thought i was the only one confused by this! this
> underscores the need for documentation.
>
> > The "context" protocol _should_ be relative to the servlet _Context_.
> > Any other approach requires a different protocol name.  I am not
> > saying that sitemap relative is not a valid approach--I am saying that
> > "context" is not a valid name for that approach.
>
> so how about this - context:// resolves to the root of the webapp,
> sitemap:// resolves to the root of the current sitemap. cocoon:// doesn't
> exist. sound good? or should we leave the confusing context:// behind us
> entirely and use webapp:// and sitemap://?

You're all right. The context:// protocol should resolve to the
environment (servlet mostly) context root. I meant that urls NOT
having a protocol specified were meant to resolve at the current sitemap
root. This was the meaning of the EntityResolver represented by a
concrete Environment implementation.

Giacomo



---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]

Reply via email to