[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Differences in Rollup
Not to be too focused on rollup, but after spending time with Paul going
back and forth on the user issue, it's occurred to me that there's also
a distinction I've been missing (and it appears others have as well),
with rollup. Alex Taler's mail earlier today highlights these in my
There's the CUA rollup which I expect can range from the extremely
simple (one set of dates) to the grotesquely complex (multiple views of
multiple input sources). I expect different implementations will
retrieve data via CAP from various sources and combine it for
presentation in a hopefully wide variety of ways. This should primarily
be used for a user to create their own personal view of calendar data.
In Alex's parlance, these would involve CALID's, the raw agenda
addresses, presumably hidden away from the user. It shouldn't be called
rollup, as from the protocol view, it's not. It's where most of the
discussion of rollup has been though.
Then there's the CS rollup, which provides a simple view of a "user".
Each implementation can figure out how to determine what sources are
used to determine this simple view. The most primitive will only allow
a single input source, while the more complex will allow multiple
sources with different ratings of how each one affects the free/busy
time. However, the choice of what gets sent out over the wire needs to
be strictly defined in CAP so that all implementations can equally use
it. This involves the use of CALADDRESSES, indicating a user, rather
than a raw calendar.
All of this is purely to solve the user demand to be able to see "is Joe
Smith busy today at 5pmEST" and potentially (if properly authorized),
the details of what he's doing.
org:Morgan Stanley Dean Witter and Discover;Information Technology
adr;quoted-printable:;;750 Seventh Ave=0D=0A;NY;NY;10019;USA