I agree with Deng.
I thought WebDAV was a reasonable choice for the central web interface
since it forced it to be resource-based and gave a bit of extra
functionality. I know that we still had it a little tangled up though.
I've no objection to abstracting it a layer away, but if it's at all
hard to add Jackrabbit on top of what remains (plugin? :) then I'd
think something has gone horribly wrong. We've (ok, you've!) already
done most of the hard work of figuring out some of the client and
locking issues - if we don't have a direct need for that then we can
document that it is only partially supported and let people that want
to fix that do so themselves, right?
Cheers,
Brett
On 13/02/2009, at 1:46 PM, Maria Odea Ching wrote:
Hi James,
I'm still undecided whether to completely remove webdav support.
From our
conversation in IRC the other day, I gleaned that we could move on
to the
new API and still support webdav right? If that would be the case,
how much
work or change would need to be made in the new API? And can we re-
use some
of the existing code in the webdav module?
Thanks,
Deng
On Wed, Feb 11, 2009 at 8:36 AM, James William Dumay <[email protected]
>wrote:
Hey guys,
I've been working on a newer repository API that is not based on a
webdav servlet but a simple HTTP servlet that supports PUT (and some
backward compatibility for Webdav's MKCOL method).
The new API has been made simple because of the limited amount of
HTTP
methods that need to be implemented.
So you may be asking "Show me the code" unfortunately I'm waiting
for an
IP waiver at Atlassian to be processed so I can get the project off
of
labs.atlassian.com and into the Archiva sandbox.
I can think of a few advantages of no longer supporting webdav:
* We no longer have to support 4 or so different implementations of
Microsoft webdav clients
* Our tenuous support for Webdav Level 3 locking can simply go away.
* The ugly issues surrounding the DavResourceFactory and the
repository
group feature will go away (currently we have two impls of a
DavResource
instead of the one)
One disadvantage I can see is that our webdav support is a
differentiator and that we already have an existing users using the
webdav feature.
Thoughts?
James
--
Maria Odea Ching
Software Engineer | Exist Global | 687-4091 | Skype:
maria.odea.ching |
www.exist.com | Innovation Delivered
--
Brett Porter
[email protected]
http://blogs.exist.com/bporter/
