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

RE: [ACL] Principal Identity

From: Lisa Dusseault <lisa@xythos.com>
Date: Tue, 20 Nov 2001 22:35:16 -0800
To: "Clemm, Geoff" <gclemm@Rational.Com>, "Webdav WG" <w3c-dist-auth@w3c.org>
The definition of dead properties is not clear enough to make that proposal
a basis for interoperability.

Since I love examples, follow me through this one:
 - A WebDAV client software package defines a couple dead properties,
"image-height" and "image-width" as integers.  It sets these properties on
all images it saves & uses them later.  Obviously so far it's a dead
 - Other clients start to use these.  Still dead, right?
 - Custom software is added onto another WebDAV server platform to set these
properties auto-magically whenever an image is uploaded.  The base server
software is uninvolved in this process.
 - A full WebDAV server decides to make sure that whenever these properties
are set, at least they're integers.  Maybe it also checks they're not
negative.  Other servers, however, still ignore these properties.
 - Another DAV server decides that it can calculate this property on the fly
and make sure it's always correct.  However, the calculation is expensive
(Note that this scenario starts to sound like justification for NOT
including properties in allprop).

When do these properties stop being dead and start being live?  Can clients
rely on these properties being listed in 'allprop' or not?  On some servers
but not on others?  Ugh.

That's why I think only a fixed list of properties can be relied upon to be
in allprop.  Let's rely on 'propname' to actually tell us the complete list
of properties, whether they're alive, dead, or somewhere in between.


> -----Original Message-----
> From: w3c-dist-auth-request@w3.org
> [mailto:w3c-dist-auth-request@w3.org]On Behalf Of Clemm, Geoff
> Sent: Wednesday, November 21, 2001 5:11 AM
> To: Webdav WG
> Subject: RE: [ACL] Principal Identity
> I believe that when we revise 2518, that that the
> new DAV:allprop behavior should be defined to be
> "all live properties defined in RFC 2518 plus all
> dead properties".  Since 2518 states that it is *all*
> properties (both live and dead), making this direction
> clear now will allow clients to rely on dead+2518
> properties.
> Cheers,
> Geoff
> -----Original Message-----
> From: Lisa Dusseault [mailto:lisa@xythos.com]
> Sent: Tuesday, November 20, 2001 1:40 AM
> To: Stefan Eissing; Jim Whitehead; Webdav WG
> Subject: RE: [ACL] Principal Identity
> > The nice thing about allprop is that you get all the dead properties.
> > The bad thing about allprop is that you cannot see if a resource is
> > versioned/version or which methods it does support.
> How do you know you can get all the dead properties?  'allprop' as defined
> in RFC2518 seems to have to change, yet we've not clearly defined a new
> behaviour.  In the meantime, I don't believe clients can rely on getting
> anything back in 'allprop' except the live properties defined in RFC2518.
> Lisa
Received on Wednesday, 21 November 2001 13:37:35 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:01:24 UTC