ISSUE-25: Weak aggregation and strong composition in containers

Aggregation

Weak aggregation and strong composition in containers

State:
CLOSED
Product:
Linked Data Platform Spec
Raised by:
Steve Battle
Opened on:
2012-10-19
Description:
Closed in very, very long discussion:

http://www.w3.org/2012/ldp/meeting/2012-11-02#ISSUE__2d_25

===

This issue pertains to questions raised by AndyS, namely how do we:

> 1/ Place an existing BPR in a container
> 2/ Place a BPR in more than one container
> 3/ Remove a BPR from a container but not delete it from the system
(c.f. 5.6.1, 5.6.2) -- this may be part of a move operation.
> 4/ Create a container under another container? Move a BPC under
another BPC?

Specifically, this raises the issue that we must be clear about how the LDP supports both weak aggregation and strong composition in containers.

The lifecycle of a composed resource is tied to the container; if the container is deleted any composed resource MAY be deleted. Weakly aggregated resources should not be deleted.

stevebattle argues that rdfs:member (and this is not atypical) defines only weak aggregation, and as such it would be incorrect to treat members as though they have a strong lifecycle dependency on the container as in: "5.6.1 If a LDPC server supports deletion of the LDPC, the server may also delete the resources that are referenced as its contents." (Linked Data Platform 1.0).

stevebattle proposes that a resource is only (strictly) composed under a container if its URI can be hierarchically relativized to the URI of the container (and without using '..' in the relativized path). Intuitively, the URI of the subordinate resource begins with that of the container.

objections include:
* URIs should be opaque
* Composition of resources in different authorities would not be possible
* Very hard to move composed resources between containers
* It would be possible to add composed resources to a container using PUT
* It would even be possible to create the container after the subordinate resource
Related Actions Items:
Related emails:
  1. LDP Rec (from eric@w3.org on 2015-02-20)
  2. Re: ldp-ISSUE-59 (recursive-delete): Reconsider usage of Aggregate/Composite construct to get predictable container delete behavior [Linked Data Platform core] (from ashok.malhotra@oracle.com on 2013-04-05)
  3. Re: A minimalist proposal for issue-50 (from pierre-antoine.champin@liris.cnrs.fr on 2013-03-16)
  4. Re: A minimalist proposal for issue-50 (from Erik.Wilde@emc.com on 2013-03-14)
  5. A minimalist proposal for issue-50 (from steve.battle@sysemia.co.uk on 2013-03-14)
  6. Re: recursive deletion (Re: Model questions) (from roger.menday@uk.fujitsu.com on 2013-01-29)
  7. Re: recursive deletion (Re: Model questions) (from henry.story@bblfish.net on 2013-01-29)
  8. Re: recursive deletion (Re: Model questions) (from ashok.malhotra@oracle.com on 2013-01-29)
  9. Re: recursive deletion (Re: Model questions) (from eric@w3.org on 2013-01-29)
  10. Re: Model questions (from sergio.fernandez@salzburgresearch.at on 2013-01-29)
  11. Re: Model questions (from henry.story@bblfish.net on 2013-01-29)
  12. recursive deletion (Re: Model questions) (from roger.menday@uk.fujitsu.com on 2013-01-29)
  13. Model questions (from ashok.malhotra@oracle.com on 2013-01-28)
  14. Re: Aggregation: simple proposal (from henry.story@bblfish.net on 2013-01-18)
  15. Re: Aggregation: simple proposal (from lehors@us.ibm.com on 2013-01-17)
  16. Re: Aggregation: simple proposal (from Erik.Wilde@emc.com on 2013-01-17)
  17. Re: Aggregation: simple proposal (from Bart_van_Leeuwen@netage.nl on 2013-01-17)
  18. Re: Aggregation: simple proposal (from lehors@us.ibm.com on 2013-01-17)
  19. Re: LDP Agenda for December 17, 2012, with a list of issues to be closed (from steve.battle@sysemia.co.uk on 2012-12-15)
  20. Re: Container membership alternative approach ACTION-27 (from steve.battle@sysemia.co.uk on 2012-12-15)
  21. Re: Container membership alternative approach ACTION-27 (from andy.seaborne@epimorphics.com on 2012-12-15)
  22. Re: Container membership alternative approach ACTION-27 (from lehors@us.ibm.com on 2012-12-15)
  23. Container membership alternative approach ACTION-27 (from Bart_van_Leeuwen@netage.nl on 2012-12-14)
  24. Approach to spec changes for resolution of ISSUE-25 Weak aggregation and strong composition in containers (from sspeiche@us.ibm.com on 2012-11-19)
  25. Re: ldp-ISSUE-30 (bugtrack): Hierarchical bugtracking service [Use Cases and Requirements] (from sspeiche@us.ibm.com on 2012-11-06)
  26. ldp-ISSUE-25 (Aggregation): Strict aggregation and strong composition in containers [Linked Data Platform core] (from sysbot+tracker@w3.org on 2012-10-19)

Related notes:

AFAIU, for reference, the initial post by Andy Seaborne was inMessage-ID: <50603364.6010607@epimorphics.com>

Olivier Berger, 2 Nov 2012, 10:35:40

Display change log ATOM feed


Chair, Staff Contact
Tracker: documentation, (configuration for this group), originally developed by Dean Jackson, is developed and maintained by the Systems Team <w3t-sys@w3.org>.
$Id: 25.html,v 1.1 2015/08/17 04:43:07 denis Exp $