Greg Stein of CollabNet is presenting. WebDAV is web-based distributed
authoring and versioning. Often just called “DAV”. This turns the web into
a writable medium that enables interoperability of web authoring tools.
It’s based on extension to HTTP, and uses XML for properties, control and
status.
RFC 2518 and 3253 are the specifications for WebDAV. The first one
didn’t include versioning because that was such a large topic on its own;
the second includes versioning.
Benefits. Document metadata is available via DAV, providing more
intelligent directories. Authors get a standard way to post content, tag it
with metadata, move and copy the content, manage versioning and prevent
inadvertent overwriting in group scenarios. Administrators perform all
interaction via the protocol; for example, you can use HTTP authentication
so no machine accounts are needed. (That’s why I use WebDAV over SSL as the
way guests can post information to my server.) If you want to reorganize
your machine’s disks, just fix the web server’s configuration to point to
the new places, and the users don’t even know it happened. WebDAV uses all
the existing web infrastructure for its own purposes.
WebDAV provides new HTTP methods: COPY, MKCOL, PROPFIND, CHECKOUT. There
are also new headers and some slight changes to existing method’s semantics.
The core functionality (RFC 2518) is pretty straightforward, versioning is
complex.
We get all the benefits of HTTP: authentication, encryption, proxy
navigation, worldwide deployment, huge talent pool, tools, applications
available.
WebDAV Clients. Open source includes cadaver (like command line
ftp client, but over WebDAV), sitecopy, Neon,
SkunkDAV, DAV Explorer, Twingle. Open source APIs exist in Python, Perl, C,
and Java. Commercial software is available from Microsoft (all of Office
and Windows Explorer), Adobe, Macromedia, and others. WebDAV is integrally
supported in Mac OS X, Windows XP, GNOME VFS, KDE “kio”, DAVFS/DAVFS2 for
Linux and Unix. Each of these make a DAV server look like a file system the
the user.
Wow! I’m using WebDAV to make all these blog postings. I’ve been
writing the files locally, then dragging and dropping the files into my
WebDAV server folder. But the speaker just said that XP fully integrates
WebDAV into the OS, not just Explorer, so any application can use it
directly. I just tried it in my GWD Text Editor (which doesn’t know
anything about WebDAV, despite it having watched me write about it)
and it worked! When I’m done with this post and save my local
copy, I can just use File/Save As to save a copy on my web server. WebDAV
servers look just like regular folders
to all applications under XP. That’s wonderful! Now I have another reason
XP is better than Windows 2000 for me (ClearType, nicer and more consistent
Wi-Fi functionality than the NIC vendor drivers give under 2000,
and a larger embedded device driver database are the others so far; those
all mostly benefit notebook users, while WebDAV is broadly useful for all
PCs).
WebDAV Servers. Apache 2.0 and 1.3 with mod_dav (written by Greg
Stein). Zope, Tomcat, Jakarta Slide, and many commercial servers. When
Stein posted mod_dav everyone was excited, and asked what they could do with
it? The answer was “well… nothing” because there were no clients. A few
clients were written by others, but when Microsoft started supporting DAV in
Windows Explorer in 1999 things started to take off.
Using WebDAV. Determine the kind of content you manage and
examine your existing workflow and users. Inventory the tools and platforms
already in use. In most cases, a WebDAV-enabled web server will be a better
file server than a regular file server because you can use all the web
functions on the server, too, for read access by others. You can also
control user roles better, too. Versioning can be supported (check out
Subversion). It even does “autoversioning” for clients that don’t know
about versioning themselves.
This has been a very useful talk for me. I’ve got to get our IT staff to
use WebDAV more.
Charles Engelke's Blog 