W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > January to March 2001

Re: Why not an encapsulation for DAV over standard HTTP 1.0 or 1.1 without required server extension ?

From: Greg Stein <gstein@lyra.org>
Date: Tue, 27 Mar 2001 03:11:33 -0800
To: Philippe Verdy <verdy_p@wanadoo.fr>
Cc: w3c-dist-auth@w3.org
Message-ID: <20010327031133.Y27539@lyra.org>
On Tue, Mar 27, 2001 at 11:47:20AM +0200, Philippe Verdy wrote:
>...
> A simple test on many web servers (despite they are using a recent Apache
> version with mod_perl extension, and even with mod_php allowed), will reveal
> that most of web servers used by web hosting providers do not allow any DAV
> protocol because of these limitation:

Okay, so you're suggesting that, say, Apache users cannot use DAV because
they don't have mod_dav installed. Fair enough, and entirely true.

Then, you go on to state that your PHP-based solution will solve the
problem. Not at all. I would still need to install that on my system.

All you've done is changed "install mod_dav" to "install PHP and this
script". That doesn't help anything.

>...
> I think personally that a light version of DAV without HTTP server extension
> is possible now, without modifying the server to enable new HTTP extensions,
> and without changing the core of DAV.

You *still* have to modify the server to enable your tunneled protocol.

> If the conceptors of DAV do not want
> to implement it, users will do it themselves !

Fine. Please feel free.

Implement it to your heart's content. I don't see Microsoft changing their
clients (Web Folders, Office, etc). I don't see Joe Orton upgrading his
clients. I doubt gnome-vfs will switch. I can't see Adobe changing all of
their clients to use it. etc etc.

DAV has widespread buy-in in its current form.

Which users do you believe *want* this kind of change? And which users will
be writing the code to make the change?

> Your answer suggests me that I could form a group to implement an
> encapsulation of DAV over POST,

You sure can. Nobody can prevent you from doing such a thing. I just don't
see that you'll get any buy-in to the approach.

The only benefit to the tunneling approach is proxy/firewall navigation. In
my experience, that has impacted relatively few people. I also believe Jim
is incorrect with his "2 years" timeline. Proxies started updating their
code about two years ago. Many proxies are now DAV-capable and are being
deployed. Again, it is rare that I see a problem report due to a proxy
problem.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/
Received on Tuesday, 27 March 2001 06:11:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:55 GMT