W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > January to March 1998

RE: WEBDAV-06-PROTOCOL Comments ( Contd..)

From: Yaron Goland <yarong@microsoft.com>
Date: Wed, 21 Jan 1998 21:27:01 -0800
Message-ID: <3FF8121C9B6DD111812100805F31FC0D0113C6B9@red-msg-59.dns.microsoft.com>
To: "'SKREDDY@us.oracle.com'" <SKREDDY@us.oracle.com>, ejw@ics.uci.edu, "'WEBDAV WG'" <w3c-dist-auth@w3.org>
12.1 - The ";" is not an error. Value: marks a BNF section. Since we don't
have a BNF production we have to put ";" in to indicate we are making a
comment.

12.2 - Fixed

12.8 - getlastmodified is based on the last modified header from HTTP which
uses the HTTP-date format. However the HTTP-date format is not the preferred
date format for XML applications, ISO 8601 is. Please refer to the mailing
archive for a detailed discussion as to why ISO 8601 was chosen. However,
for backwards compatibility reasons we have chosen to keep HTTP-date for
last modified while moving to ISO 8601 for the brand new  property,
creationdate.

Defining Author/Owner - If the author/owner property is needed for ACLs
(which I don't believe it is, ACL ownership can always be implicit based on
creator or the administrator's whims) then it will be defined by the ACL
effort. Please refer to  the IETF website to read the latest draft of the
ACL requirements and protocol documents. The ACL effort is currently quiet
because the working group is focused on the main document last call. As soon
as the last call is over you can expect the ACL debate to start up. Again, I
would direct you to the mail archive for records of the ACL debates that
have happened to date. I would especially recommend that you read the last
author's meeting summary, it contains a long of good information on the open
issues, of which there are many.

Annotations - When DAV first started it looked into the possibility of
defining all sorts of useful properties and quickly realized it was slipping
into a bottomless pit. Since we were already over our quota for bottomless
pits the working group decided that a property would ONLY be defined if it
was absolutely required for the proper functioning of the protocol. Since
annotations do not go under that heading, they would not be included. That
doesn't mean that the property isn't useful, just that one can profitably
use the DAV protocol without it. We depend upon efforts such as z3950 and
Dublin Core to define such terms as author and annotations. It turns out
that their definitions are non-trivial. BTW if your curious how display name
and creationdate made it through the answer is that everyone both needed and
agreed on the definition of display name. Creationdate is an interesting
question. I'm not sure how it got in. I know that we really need it but that
doesn't necessarily mean it should be in the protocol. Jim, do you remember
how creationdate got in?

			Yaron

PS Just in case there is a misunderstanding, new properties can be defined
at any time. In the worse case old servers will just treat them as being
"dead", meaning that it will just record their values but not enforce their
syntax or semantics.


> -----Original Message-----
> From:	Surendra Reddy [SMTP:SKREDDY@us.oracle.com]
> Sent:	Wednesday, January 21, 1998 4:32 PM
> To:	ejw@ics.uci.edu
> Subject:	WEBDAV-06-PROTOCOL Comments ( Contd..)
> 
>   Sorry for sending these comments in batches:
> 
>  Here are some more corrections required in the document:
> 
> 12 DAV Properties:
> 
> 	Section 12.1
> 
> 		Remove ";" before "The time and date ...." or provide a
> missing value
> identifier like ISO8601-date
> 
> 	Section 12.2
> 		
> 		Purpose:	   "Provies ...."  change to "Provides ...."
> ( typo )
> 
> 	Section 12.8
> 
> 		getlastmodified
> 
> 		Value:	HTTP-date ; defined in section 3.3.1 of [....]
> 
> 			rephrase this as same as  12.1 value. 
> 		Note: 	creationdate and lastmodified properties should have
> the same value
> types. Correct me if I am 
> 			wrong.
> 
> NOTE:	Do we have a property that retrieves the author/owner of the
> document.
> Unless we define document ownership, how does ACL
> 	is going to work? Also,  how do we get  "annotations of the
> document"?
> Don't we need a property that defines document
> 	annotations.
> 	
> 	Can we add these TWO properties? Any comments or suggestions?
> 	
> 
>    
>    Regards,
>    -Surendra
> 
>   ------------------------------------------------------ 	
> Surendra K Reddy         Tel.   +1(650) 506-5441 
>                          Fax.   +1(650) 506-7421
>                          Email.	skreddy@us.oracle.com
> "It is unwise to be too sure of one's own wisdom. It is healthy to be
> reminded
>  that the strongest might weaken and the wisest might err."  
> 
>   
> Warning: Statements and opinions stated herein may not be those of Oracle
> Corp.
>   
>  << File: ATT332208.ATT >> 
Received on Thursday, 22 January 1998 00:27:21 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:44 GMT