Item clouds for semantic item aggregrations

October 26th, 2004 at 6:49 pm (3 years, 8 months ago) by Andi Vajda under chandlerdb

Item clouds are a way of semantically grouping together items of different kinds not otherwise stored together in a ref collection or otherwise. For example, the constitutive parts of an email in Chandler are items of various kinds such as the headers, the body, the attachments etc… The email item cloud is defined to be that set of constitutive items. When copying an email one really means to copy all of these items. The same applies when sharing it or with any other high level operation meant to be used with the high level email object.

Morgen describes clouds in his journal entry in greater detail.

As with ref collection aliases, the cloud feature started small and became more mature as its usage scenarios became more evident. For example, originally the repository was implementing attribute-specific copy policies specifying if an how an attribute’s value should be copied during an item-copy operation. The main flaw with this approach is that such a copy policy is set for the attribute and is global for all the items using the attribute and for all possible copy scenarios. We decided to move away from attribute-specific copy policies and move to cloud-based copying instead. One or more clouds are defined for a kind and one is chosen for a particular operation - such as copying or sharing an item - based on the needs and context of the operation. One could imagine different clouds being used just for sharing based on the level of trust between the sharer and the sharee.

It is expected that item deletion is also going to follow this trend and move away from attribute-specific delete policies as is currently implemented.

Cloud and cloud endpoint inheritance are also somewhat unusual and are described in Morgen’s document.

Share and Enjoy: These icons link to social bookmarking sites where readers can share and discover new web pages.
  • Digg
  • Sphinn
  • del.icio.us
  • Facebook
  • Mixx
  • Google
  • Reddit

Leave a Reply