Mailing-List: contact jackrabbit-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: jackrabbit-dev@incubator.apache.org
Received-SPF: pass (asf.osuosl.org: local policy)
From: "Daniel Florey" <daniel.florey@web.de>
To: <jackrabbit-dev@incubator.apache.org>
Subject: AW: webdav server(s)
Date: Thu, 15 Dec 2005 13:18:01 +0100
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <43A0720E.1030806@ipov.net>
Thread-Index: AcYA5I8ZRvBDonfORrO6PYvOp99uhgAiKCLw
Message-Id: <E1Ems36-0006K6-00@smtp06.web.de>
Sender: daniel.florey@web.de

Hi folks,
this in an interesting thread.
It would be great to have a full featured WebDAV-stack implemented in =
Java
with a simple and consistent api.
When the Jackrabbit project started, I was little bit frustrated that =
people
started to reinvent the wheel instead of using simple
WebDAV-to-java-method-mapping as api. So I stuck with the Slide project =
even
though the codebase is really a mess and nobody really likes to work on =
it.
If there would be a combined effort to implement a robust WebDAV-stack, =
I'd
really appreciate this.
Let me summarize the goals of such a project from my point of view:
1. The big advantage of WebDAV is that users can (in theory) use =
existing
clients to communicate with the server. There are many clients =
supporting
the very basic WebDAV set, but not many using any advanced =
WebDAV-features
(Delta-V, DASL, ACL).=20
The best WebDAV-clients supporting Versioning (and this seems to be the
hardest part to implement) are subclipse and TortoiseSVN.
So my primary goal of an apache WebDAV-project would be to be =
interoperable
with subversion. This could also include close communication with the
subversion team and perhaps changes to make subversion more =
WebDAV-compliant
and discussions on the DeltaV-mailing list to incorporate extensions =
that
subversion introduced into the spec if necessary.
It could be part of the WebDAV-project to provide new features for the
subclipse-Plugin to support ACL, DASL, Bindings and other =
WebDAV-features
right out of eclipse.
This could lead to a RCP-solution to provide a robust and mature
cross-platform WebDAF-client.
2. Support of all important specs that you can find on Julian Reschke's
great webdav site:
http://www.greenbytes.com/tech/webdav=20
3. The server implementation should be open to make it possible to add
support specs easily. This requires a good concept of virtual resources =
and
live/dead properties. But theses details can be discussed later on.
4. Provide a single api for both local and remote repository =
communication.
In case of remote repository communication of course WebDAV should be =
used
as a protocol to enable communication with existing WebDAV-servers (Ms
Exchange, SAP Netweaver, Software AG's Tamino-Server etc).
5. Support for feature examination. The api should reflect the different
levels of WebDAV support that the server provides. ACL's can be exposed =
in
different ways, some specs may not be supported by certain servers. This
should be reflected in the api.

We are currently discussing to move Slide to TLP at the
Slide-dev-Mailinglist. Personally I'd prefer to create a new
webdav.apache.org TLP and start something more elegant from scratch. If
there is a chance to combine the efforts taken in the Slide and in the
Jackrabbit project concerning the WebDAV support, I'd be glad to support
this effort.

Cheers,
Daniel


> -----Urspr=FCngliche Nachricht-----
> Von: =
jackrabbit-dev-return-4903-daniel.florey=3Dweb.de@incubator.apache.org
> [mailto:jackrabbit-dev-return-4903-
> daniel.florey=3Dweb.de@incubator.apache.org] Im Auftrag von Robert r.
> Sanders
> Gesendet: Mittwoch, 14. Dezember 2005 20:27
> An: jackrabbit-dev@incubator.apache.org
> Betreff: Re: webdav server(s)
>=20
> As someone who has been monitoring both Jackrabbit and Slide lists =
this
> idea struck me before.  If the people still working on Slide could be
> brought over to work with some of the Jackrabbit people it seems to me
> that this would be in the best interests of both communities.  I don't
> know how this would be done, or wether a TLP or a subproject would be =
a
> better alternative. While they aren't the same (and I am not an expert
> on either) it seems obvious that there are some gains to be made by
> having a full featured WebDAV server with a JCR backend.
>=20
> St=E9phane Croisier wrote:
> >
> > There is a thread running on right now on the Slide Dev List about
> > moving Slide to a Top Level Apache Project (TLP). Currently we are
> > using Slide as a temporary solution to store some binary files for =
our
> > CMS and I agree quite a lot with the conclusions of Brian... The =
Slide
> > project is nearly dead and without nearly any activities from
> > beginning of this year.
> >
> > So the best solution would certainly be to open a new =
"dav.apache.org"
> > top level project which would include a Slide 3.0 full refactoring
> > based on Jackrabbit (or whatever would be the new name of such a
> > project) and try to gather all the currently scarce DAV ressources =
and
> > expertises into one single project.... Then CalDAV or other DAV
> > extensions would then also easily fit in such a new TLP...
> >
> > My 2 cts...
> > St=E9phane
> >
> > At 17:31 14.12.2005, Brian Moseley wrote:
> >> Stefano Mazzocchi wrote:
> >>
> >>>> personally, i have no interest in working on such a thing. i
> >>>> wouldn't tell somebody not to do it, but i wouldn't help them =
either.
> >>> Can you tell us why?
> >>
> >> i found the slide codebase to be extremely confusing, verging on
> >> incomprehensible. i was unable to make heads or tails of the apis =
and
> >> had only the vaguest inkling of how i might extend it for caldav.
> >>
> >> by contrast, the jcr-server design is relatively simple and =
elegant,
> >> and the extension points are natural and obvious.
> >>
> >> also the slide community didn't seem to have much momentum back in
> >> the spring of 2005. there was no defined release plan and extremely
> >> little support on the mailing list. the documentation that existed
> >> was sparse and often frustratingly unintelligible, so when you had
> >> questions, you were basically screwed.
> >>
> >> things might have changed for the better with slide, but i'm not
> >> optimistic. i vastly prefer the jackrabbit code and community.
> >
> > - -- --- -----=3D[ scroisier2 at jahia dot com ]=3D---- --- -- -
> > www.jahia.org : The Java Unified Web Platform
> >
>=20
> --
>     Robert r. Sanders
>     Chief Technologist
>     iPOV
>     (334) 821-5412
>     www.ipov.net