Re: Does anyone still think that RECURRENCE-ID's do NOT change?

Thanks for the input.
We've been hashing this out with Doug for about 8 weeks now.

If you go through the archives over the past 2 months
almost every message has been on this topic.

Your arguments are very reasonable, and unfortunately
reason has not been very useful here.  Doug clings to
his ideas like they were his saving grace and ignores
any issues raised against his idealogy that he can't
disprove.  He also twists it said to mean what he wants
it to mean rather than the its intent and focuses on
stupid irrelevant arguments without addressing the
core issues.

Your "2 cents gut feel" is actually very demonstratable
to be the only sane way and there have been megabytes
of posts to this list cementing how not doing it the
"fixed-id" way is extremely fragile.

At the moment Doug Royer is the only proponent of his
model and he has placed me on his ignore list because
I demanded he prove his model didn't break.

If you do want to participate, your best bet is to talk to
Doug directly, the rest of us already agree on what we're
saying though we say it in very different ways sometimes.

8 weeks and probably some 300 messages is an eye opening
eperience in how rigomortized some people can be.  We
even sought the opinion of the original authors who
unfortunately didn't do a crystal clear enough job to
convince Doug that he's still mistaken.

-- Michael --


"Joerg Reichelt" <[email protected]> wrote in message
news:CA8D7A11C649284AA163D951DA56533001305477@srv-mailsjc2...
>
> It appears that it would have been much easier - and a lot less
> controversial - to define the recurrence ID as an index into the
> set of instances.
>
> Looking at it this way, it would be clear to most of us that
> - this index DOES NOT change for simple changes to that instance
>   it refers to
> - this index DOES change when the instances of a set need to be
>   re-indexed, e.g. when the definition of what instances belong
>   to the set changes; when this happens, and you want to keep
>   modifications to individual instances of the set, you need to
>   update the index in each of the descriptions of the instance
>   modifications (e.g. change that index to the new index of the
>   modified instance)
>
> Since the recurrence ID is currently defined to be constructed of
> the start date and time of the instance, it appears to imply that
> it always would be equivalent to that value, which seems to be the
> source of the disagreement. The argument is revolving around the
> question whether the recurrence ID should be the original or the
> current value for a given instance (and conflicting references to
> RFC paragraphs about what is correct have been presented, and both
> alternatives have been extensively discussed).
>
> It appears to me that it would be easier to work with a constant
> reference as much as possible than with one that changes all the
> time, but that is just my personal opinion on this topic.
>
> My 2 cents.
>
> - Joerg
>