Re: iCalendar as an XML DTD

["Alex Hopmann" <hopmann@xxxxxxxxxxx>]

I just wanted to respond to a few comments and dispell a few misconceptions.

First of all the comparison to ASN.1 and several statements made about the
complexity of XML are unfounded. XML has (to me) clearly passed the standard
"IETF Implementation" test in that there are literally dozens of
implementations, in a variety of environments from PERL, to Java to C++.
Most importantly these implementations have for the most part been shown to
interoperate, and creating new ones isn't rocket science- I would
characterize it as "a weekend project for a graduate student", which is the
measure we have often discussed in IETF working groups for this sort of
thing.

Part of the strength of XML is that it is a fairly simple text based format
that can represent different types of data. Personally I find that XML can
create much more simple representations of things that some alternative
formats. The iCalendar spec has plenty of examples- Much of the complexity
of iCalendar involves efforts to get around the lack of true hierarchical
representation of properties/sub-properites in the underlying MIMEDIR
format. As it is, it can be pretty hard to put iCal data into a database and
then construct useful queries on it.

Having said that, I don't see any crisis. Long term I think an XML version
of iCalendar is pretty much inevitable. Short term, we are doing useful,
productive work finishing up iCalendar now. If we do our jobs right, any
content conversions we might need in the future should be easy. While
multiple representations of a data object can be a pain, in fact it is the
use of XML as a common representation/serialization mechanism for many types
of objects that is its greatest potential.

Alex Hopmann
Microsoft Corporation