To what extent is the SSE history used in the sync framework RRS feed

  • Question

  • Hi,

    Although the support of SSE compatible providers is mentioned, there seems to be little information in that area. I would like to know how the Sync Framework uses the SSE, specifically whether the full history for an item is required for correct results. We are considering to limiting the SSE history to contain only the originator/time stamp for the last update.




    Tuesday, November 20, 2007 12:55 PM


  • Hi Tudor,


    In order to interop with SSE Microsoft Sync Framework translates SSE history for items in a feed into the sync knowledge and vice versa. This allows Microsoft Sync Framework to see SSE endpoints equal to the Microsoft Sync Framework endpoints in the sync community. You can read more about the sync knowledge in my blog. During conversion from SSE history to the sync knowledge, Microsoft Sync Framework produces clock vector elements for each latest SSE history entry from each unique originator ("by" attribute value). This procedure "trims the fat" and leaves only latest SSE history entries in the sync knowledge, while preserving the correctness of the sync metadata. Later the produced sync knowledge is treated just like any other sync knowledge. During the reverse translation, the SSE history for a given itemis produced during sync knowledge. Based on the above, as long as you keep the latest SSE history entry (you have to do so anyway according to the SSE spec), you should be able to do multi-master sync in the community which includes SSE endpoints.

    Note that Microsoft Sync Framework also includes fairly high level APIs which allow you to produce and consume RSS or Atom feeds without writing SSE sync provider.



    Dev in Microsoft Sync Framework

    Tuesday, November 20, 2007 6:58 PM