Re: A question on Versioning-unaware cl

Chris Kaler (ckaler@Exchange.Microsoft.com)
Tue, 12 Oct 1999 10:33:51 -0700 

Message-ID: <FD7A762E588AD211A7BC00805FFEA54B041DD983@HYDRANT>
From: "Chris Kaler (Exchange)" <ckaler@Exchange.Microsoft.com>
To: "'jamsden@us.ibm.com'" <jamsden@us.ibm.com>, ietf-dav-versioning@w3.org
Date: Tue, 12 Oct 1999 10:33:51 -0700
Subject: RE: A question on Versioning-unaware cl


Many protocol documents have implementation suggestions.
These are things that the design team thought of just in
case other people haven't.  Remember, the design team tends
to think through the problem more deeply that a lot of
implementors.

I disagree with the blanket statement that SHOULD is a
candidate for removal.  There are always areas that cannot
be black or white and need to be addressed, if for no other
reason than completeness.

Chris

-----Original Message-----
From: jamsden@us.ibm.com [mailto:jamsden@us.ibm.com]
Sent: Tuesday, October 12, 1999 10:25 AM
To: ietf-dav-versioning@w3.org
Subject: RE: A question on Versioning-unaware cl




This looks like an area we might want to stay out of. If the protocol wants
to
say something about implementations, then there might be something wrong
with
the protocol. It also doesn't seem necessary for us to tell implementors
that
they should maintain the integrity of their underlying repositories, or how
they
should do it. Finally, anything with SHOULD is a candidate for removal from
the
spec.





"Chris Kaler (Exchange)" <ckaler@Exchange.Microsoft.com> on 10/12/99
12:53:41 PM

To:   "'Tim_Ellison@oti.com'" <Tim_Ellison@oti.com>,
ietf-dav-versioning@w3.org,
      masinter@parc.xerox.com
cc:

Subject:  RE: A question on Versioning-unaware cl



I think all we need to do is have an implementors note that
says if you use other protocols to access your store, and they
don't support versioning, e.g. FTP, then updates from those
protocols SHOULD support automatic versioning to maintain
integrity in the repository.

Chris

-----Original Message-----
From: Tim_Ellison@oti.com [mailto:Tim_Ellison@oti.com]
Sent: Tuesday, October 12, 1999 6:54 AM
To: ietf-dav-versioning@w3.org; masinter@parc.xerox.com
Subject: RE: A question on Versioning-unaware cl



Larry wrote:

># I admit, that they typically share the same files on a disk and thereby
># share data, so you can 'put' with FTP and 'get' with HTTP, but that's not
># the protocols interoperating, that's merely a shared disk.
>
>http and ftp 'interoperate' because their URL schemes were designed
>to allow possible substitution of the scheme name while leaving the
>rest of the URL alone. The design decision to try to align the URL
>schemes was intentional and discussed at length. It wasn't really
>very successful: the cases where it is possible to do the substitution
>are rare.

.. and getting rarer as people use active server pages, JSP's, server side
includes, etc. that usually only come as part of HTTP server behavior.
 Nowadays what you get via HTTP is unlikely to be what you get via FTP.

>If ftp didn't use the same hierarchical syntax as http, they wouldn't
>interoperate.
>
>Since WebDAV uses the same URL scheme as HTTP for its resources, it
>is reasonable to believe that FTP would continue to interoperate
>with WebDAV, if you 'put' with HTTP or webdav and 'get' with FTP.
>
>Whether or not ftp/webdav interoperability is "in scope" for the
>deltaV working group, it would be useful to at least note this
>issue as implementation advice.

What advice are you proposing?  The URL formats are being well defined by
WebDAV.

>There was some hope that NFS (webnfs) and http would interoperate,
>too. A mapping between WebDAV and NFS v4 would be as interesting
>as the mapping with FTP.