Re: ADD does not INVALIDATE current PATTERN
"Michael Fair" <[email protected]>
| Newsgroups | gmane.ietf.calendar |
|---|---|
| Organization | DaClubhouse |
| Message-ID | <[email protected]> |
> I agrees that it changes the set, does not change any RECURRENCE-ID on > existing set. I am trying to avoid the word CHANGE PATTERN, because it > has been said that any change in PATTERN would cause all RECURRENCE-IDs to > be invalid. If you can word smith this better please do so... Thanks I'd have to say that in many respects I agree with the assement. In the same way a new SEQUENCE value always obsoletes/invalidates the old values every pattern change will also cause a SEQUENCE bump. Now most often there will be many RECURRENCE-IDs that exist in both the old and the new sets. Existence in both sets is really meaningless because the SEQUENCE increase invalidates the attendence STATUS. However, assuming that the instance hasn't changed (e.g. if there were any instance updates those updates were propogated), then it's safe to just automatically transfer the old STATUS to the new SEQUENCE version of the object. Perhaps something along the lines of: "ADD can only add to the number of instances in the current pattern." Or, if you're looking for something about the pattern change: Any pattern change that results in the exclusion of any prior RECURRENCE-ID value implicitly cancels any instances referenced by the now excluded values. This implies that a CUA discovering that it has now invalid instances should remove those and only those instances whose RECURRENCE-ID is now excluded. -- Michael --