Mailing-List: contact users-help@jackrabbit.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@jackrabbit.apache.org
Received-SPF: pass (athena.apache.org: domain of aklimets@day.com designates
 207.126.148.183 as permitted sender)
Message-ID: <c3ac3bad0810260724g4d4306e1p4fc40a1d12625d81@mail.gmail.com>
Date: Sun, 26 Oct 2008 16:24:22 +0200
From: "Alexander Klimetschek" <aklimets@day.com>
To: users@jackrabbit.apache.org
Subject: Re: WebDAV Paradigm Mismatch
In-Reply-To: <20173656.post@talk.nabble.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
References: <20173656.post@talk.nabble.com>

Sorry, but your rant is a bit unclear: what exactly are the issues you
face with WebDAV on Jackrabbit? Let me clarify on the various issues
you threw in:

Files and folders: The major WebDAV on Jackrabbit case (provided by
the SimpleWebdavServlet under /repository in the jackrabbit webapp
[1]) is mapping files and folders onto nt:file and nt:folder nodes in
JCR. That way all you see and can do with the WebDAV on Jackrabbit are
natural files and folders.

Run out-of-the-box: Jackrabbit today is a "backend" library
implementing the JCR spec (it is the reference implementation) and it
is to be used in other (web) applications. Thus there is currently no
goal to have a one-click-installer for end-users on all platforms to
provide a generic document storage. There has been work on such a
thing in the Apache Sling project, which provides a runnable jar with
Jackrabbit included [2], that after startup, gives you a WebDAV on
http://localhost:8080 (although it has a slightly different mapping to
JCR nodes), and a commercial JCR implementation partly based on
Jackrabbit has something called "quickstart", ie. a runnable jar with
Jackrabbit's WebDAV out-of-the-box at
http://localhost:port/repository/<workspace>/ [3].

Locking/Versioning: Do you want to version the files after every save?
That is not done by default since it would create many unnecessary
versions (nothing you could generically assume for all files), but it
can IMHO be quite easily customized by writing your own variant of the
SimpleWebdavServlet.

[1] http://jackrabbit.apache.org/jackrabbit-web-application.html
[2] http://incubator.apache.org/sling/site/downloads.cgi
[3] http://dev.day.com/microsling/content/blogs/main/firststeps1.html

Regards,
Alex

On Sun, Oct 26, 2008 at 3:52 PM, Mike Oliver <moliver@corenttech.com> wrote:
>
> As a committer on the Slide project before its retirement, I have a number of
> applications built on top of it and some of those are better on JackRabbit
> due to the Node Types and other aspects not relevant to this post.
>
> But WebDAV on JackRabbit sucks.  The reason it sucks is that the WebDAV
> filesystem paradigm is what users, especially non computer scientist users
> expect and using WebDAV by going to My NetWork Places and adding a WebDAV
> folder needs to show Folders and Files.  With Slide this works Out of the
> Box and you can still download and install the Slide Tomcat bundle and start
> using it with WebDAV in a few minutes and use Microsoft Word to save a
> document to a Web Folder, and as long as Word has the document open, it is
> locked in Slide for all to see, Save it after some changes and it stays
> locked but you get a new version.  Close Word and the lock goes away.  You
> can have folders within folders and documents within folders.
>
> Simple, even my wife can use it.
>
> I am tempted to take Slide, create a new set of Stores for Slide that
> implement JackRabbit as the repository and use the Slide WebDAV interface.
>
> As I said for some things JackRabbit is greatly superior to Slide, but for
> WebDAV and Web Folders...uh no.
>
> Ollie
> --
> View this message in context: http://www.nabble.com/WebDAV-Paradigm-Mismatch-tp20173656p20173656.html
> Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
>
>


-- 
Alexander Klimetschek
alexander.klimetschek@day.com