W3C home > Mailing lists > Public > www-webdav-dasl@w3.org > April to June 1999

RE: use of DAV namespace

From: Yaron Goland <yarong@microsoft.com>
Date: Mon, 17 May 1999 17:35:44 -0700
Message-ID: <3FF8121C9B6DD111812100805F31FC0D087931F7@RED-MSG-59>
To: "Lisa Lippert (Dusseault) (Exchange)" <lisal@exchange.microsoft.com>, Jim Davis <jdavis@coursenet.com>, ietf-dav-versioning@w3.org, www-webdav-dasl@w3.org
If the meanings of the names are not 100% identical then two completely
different names must be used. How this is resolved, either through
coordination between DASL and Delta-V or through the use of two subsetted
namespaces is of less importance than the central principal that identical
names must have identical meanings.

<pedantic ramblings>
If we do decide to split the namespaces I would like to see us use the
format dav:v and dav:d (the bytes you save may be your own) rather than
dav://v/ and dav://d/. I would be extremely concerned were we to give our
users the mistaken impression that the DAV namespace is some hierarchy with
meaningful names. That is, if you get dav://v/foo you can say that foo is
necessarily a versioning feature. This would be a mistake as it overloads
the semantics of the namespace. All dav names should be treated as
absolutely flat and the only valid comparison function should be a
byte-by-byte compare. I think using the dav:v and dav:d format make this
point clearer and reduces the chance that someone will mistakenly think that
dav:v:foo has any special meaning.

Of course, on the negative side, everyone will end up mistyping dav://v and
we will all end up programming our DAV parsers to accept dav://v but damnit
at least we will know in our hearts that we were right! =)
</pedantic ramblings>

			Yaron

> -----Original Message-----
> From: lisal [mailto:lisal@microsoft.com]
> Sent: Mon, May 17, 1999 4:21 PM
> To: Jim Davis; ietf-dav-versioning@w3.org; www-webdav-dasl@w3.org
> Subject: RE: use of DAV namespace
> 
> 
> I don't see a drop-dead problem with this, if we're going to 
> allow, as you say, that WebDAV XML cannot in general be validated.
> 
> The DASL use of DAV:and will be within DASL containers like 
> <DAV:basicsearch> (or is it simplesearch)?  Versioning will 
> use it within DAV:basicrsr I think.  There is sufficient 
> context for it to be clear which is meant.
> 
> Actually, it may be even better than that, if we're careful:  
> if we can make sure the meaning of the versioning DAV:and 
> consistent with the DASL DAV:and.
> 
> In general, you point out a really useful consideration, 
> which is to be aware of what XML properties all other DAV WGs 
> are defining.  It's too late to wait until they're proposed 
> standards, when it's so easy to deal with it earlier.
> 
> Lisa Lippert
> 
> > -----Original Message-----
> > From: Jim Davis [mailto:jdavis@coursenet.com]
> > Sent: Monday, May 17, 1999 4:02 PM
> > To: ietf-dav-versioning@w3.org; www-webdav-dasl@w3.org
> > Subject: use of DAV namespace
> > 
> > 
> > The Versioning Extensions (Kaler et al, Jan 20, 1999) draft 
> defines a
> > number of XML elements (some are properties, others just 
> > elements) , all in
> > the DAV: namespace.  Of these, at least three conflict with elements
> > defined by DASL.  (DAV:and, DAV:or, and  DAV:not). 
> > 
> > Thus It would not be possible to construct a single DTD for 
> > WebDAV with
> > both versioning and DASL.  I understand that, in general, 
> > WebDAV XML can't
> > be validated (because we allow undefined elements to be used, for
> > extensibility), still it seems bad if two different DAV 
> extensions are
> > incompatible.
> > 
> > I suggest that both extensions (DeltaV and DASL) used a new 
> > namespace, at
> > least for those elements they introduce.
> > 
> > Would it be possible to use DAV://versioning/ and DAV://dasl/ 
> > respectively?
> > 
> > Besides the three conflicting elements, there are a number of 
> > others that
> > seem to me to have rather "generic" names, that is, I could 
> > imagine other
> > DAV extensions that might want to use these names.  I 
> > understand that tag
> > element names are not user visible, but still for the sake of 
> > programmers
> > it is useful to have meaningful names.  To avoid future 
> > confusion, it might
> > be better to put all new versioning elements in a new namespace.
> > 
> > The potentially conflicting elements names include:
> > 
> > DAV:comment
> > DAV:report
> > DAV:basetime
> > DAV:inheritancetype
> > 
> > I've only just joined the deltaV list, sorry if this has 
> > already come up.
> > 
> > regards
> > 
> > Jim Davis
> > 
> > please reply to jrd3@alum.mit.edu, despite the Reply-To 
> > address in the header.
> > 
> 
Received on Monday, 17 May 1999 20:35:51 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 22 March 2009 03:38:04 GMT