KML output

[Hal Mueller]

I attended the BoF, and also spoke quickly with one of the Google
gaggle the next day.

My take on the BoF was that we should expect nothing until after the
first of the year; that one primary reason for Google's libKML is to
solidify, via example, some sloppiness in the KML spec; and that we
will see a fairly complete Google-authored libKML released from them
before it is opened up for community participation.

The guy I spoke to on Thursday (not James MacGill, one of the other
two) said that he expected the first draft would provide a round-trip
capability:  read a KML file, write the resulting object back out to
KML, with presumably identical input and output.

I think that it's only around the edge cases of the KML spec where we
might see differences between a community-developed libKML and a
Google libKML.

Hal Mueller