jakarta-taglibs-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dan Shields" <dan...@shields.org>
Subject Re: [PROPOSAL] WebDAV tag library
Date Thu, 06 Mar 2003 09:08:43 GMT
Just to be perfectly clear, the test case I am trying to satisfy is:

<dav:putResource url=http://webdav.webfolders.com/workspace">
  <cvs:checkout repository=":ssh:user@repo.repository.com:/repository"
module="my-module"/>
</dav:putResource>

The cvs:checkout element will result in a series of putResource callbacks.
You'll notice that the life-cycle of the dav:putResource element entirely
surrounds the life-cycle of embedded cvs: element.

----- Original Message -----
From: "Willie Vu" <ctwillie@ust.hk>
To: "'Tag Libraries Developers List'" <taglibs-dev@jakarta.apache.org>
Sent: Tuesday, March 04, 2003 3:39 AM
Subject: RE: [PROPOSAL] WebDAV tag library


Dan,

Thank you for your feedback.

> -----Original Message-----
> From: Dan Shields [mailto:daniel@shields.org]
> Sent: Friday, February 28, 2003 12:32
> To: Tag Libraries Developers List
> Subject: Re: [PROPOSAL] WebDAV tag library
>
>
> Hi Willie & Pierre,
>
> Great proposal Willie, I agree with it wholeheartedly.  I
> also enjoyed Pierre's questions, which inspired me to throw
> in some of my own personal experience.
>
> I visited the potential use-cases (or lack thereof) for a
> WebDAV taglib, recently, through the development of my own
> tag library extension, one that spoke the CVS protocol to a
> remote CVS server.  The problem with this idea was much the
> same, that is: adding CVS protocol to a JSP page would bring
> much potential complexity, and how could I showcase the
> usefulness of this feature?  Most of concern: where would the
> files be written and read from
> for operations such as checkout, update, status, etc.?   What
> would be the
> use?   It's difficult to fathom, on it's own...
>
> However, I think it would be real nice if I could perform a
> checkout operation from a CVS repository and have the result
> written directly to a web-accessible workspace without
> requiring intermediate storage.
>
> I observe that both a CVS taglib and a WebDAV taglib,
> considered alone, suffer from a common challenge:  they move
> data around that strains at the confines of simple EL variables.
>

I agrees with you.  JSP is designed to handle character streams.  That's why
the
WebDAV taglib has to rely on a ResourceContentServlet that handles upload
and
download.

> Let it be resolved that:
> I want my tag libraries to talk directly to the public
> interfaces of other tag libraries with which I seek to
> exchange datasets.
>
> This technique is already present in core, io, standard, and
> xtags, using a commonsense public interface approach, but I
> think that the sandbox could benefit from seeing more of this
> in linking together focussed sub-projects.
>

I am still unclear how this technique resolves the above problem.  By any
means,
users using web-based WebDAV client has to upload and download resources to
their local PCs.


> My projects will soon reside in the taglib sandbox as
> actively maintained sub-projects, once the architecture is
> solidified.  What do you say?

Can you give me more info on what you are attempting to do?

--

Willie Vu


---------------------------------------------------------------------
To unsubscribe, e-mail: taglibs-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: taglibs-dev-help@jakarta.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: taglibs-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: taglibs-dev-help@jakarta.apache.org


Mime
View raw message