- From: Judith Slein <slein@wrc.xerox.com>
- Date: Thu, 25 Jun 1998 13:01:50 PDT
- To: w3c-dist-auth@w3.org
The following changes were agreed upon by the group at Redmond. Unless I hear protests from the mailing list, I will make these changes: 1. The "internal member" (directly contained) terminology will be removed. References should be contrasted with ordinary resources that have content, not with internal collection members. 2. Clarify requirement 3.1.5 by inserting the word "target", so that the requirement reads: "Operations on a target resource do not affect references to it except as needed to enforce referential integrity." 3. State explicitly that the requirements are silent on how the server maintains referential integrity. 4. State explicitly that there is no requirement that references be acyclic. 5. Review the requirements to see whether it makes sense to separate the notion of referencing from the notion of collections. If so, restate all referencing requirements so that they are independent of collections. (I have reviewed the requirements with this in mind. There are a few that are specific to the context of collections, but for the most part they are independent of collections. I will rephrase them to talk about "references" or "referential resources" instead of "referential members of collections" except in the places where collection membership is really integral to the requirement.) (2 and 3 assume that strong references will stay in the requirements. If not, 2 can be simplified and 3 can be dropped.) Ordered Collection Requirements -- No Changes Ordered collection requirements were relatively uncontroversial. The restriction to a single ordering per collection was revisited and confirmed. If additional orderings are desired, they can be achieved by created separate collections with those orderings (though this would be a maintenance headache) or on the client side. The suggestion that there be a requirement relating to scalability and performance was rejected on the grounds that this is motherhood and can be assumed for any protocol specification, and that metrics are too difficult to define. --Judy Name: Judith A. Slein E-Mail: slein@wrc.xerox.com Phone: (716) 422-5169 Fax: (716) 422-2938 Xerox Corporation Mail Stop 105-50C 800 Phillips Road Webster, NY 14580
Received on Thursday, 25 June 1998 15:56:06 UTC