[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [higgins-dev] IdAS: propose moving applyUpdates
|
This sounds okay to me, every CP will be faced with potential interesting ways of handling applyUpdates at every level. I'll add an IDigitalSubjectContainer to the Basic notification impls I'm doing.
And we'll leave the nucularity up to ol' dubya to pass a bill stating that all CPs are entitled to pursue it in accordance to the dictates of their own conscience!
Tom
>>> "Jim Sermersheim" <jimse@xxxxxxxxxx> 5/2/2007 8:58 PM >>>
In today's F2F mtg, we agreed that metadata should be able to exist on an IContext. Of course, this leads to the question of how it's updated. Just like elements below a digital subject, the user may need to perform a number of updates and then signal to the CP that the state is appropriate for application to a backing store. Not only this, but there may also be CP's with backing stores which need this kind of signal when multiple digital subjects are being updated.
Probably the best way to handle this is to move applyUpdates() up to IContext. We would keep the same semantics we have today, only the grouping would be up a level. Meaning applyUpdates applies the updates made thus far to member data (metadata and digital subjects) held by the context and to any updated sub-elements of those. As before, atomicity (oops, Mike wants me to use the word nucular since I'm in Texas, so -- nucularityishness) is CP-dependent.
Jim
