Draft minutes now available for comment

[Syd Bauman]

| Combining characters such as ligatures should not be represented
| as nested <c> elements: the content of a <c> might be multiple
| codepoints.

Indeed. In P4 the <c> element cannot nest (its content is #PCDATA
only); is there any reason to do otherwise in P5?

<p>| Definition of a TEI namespace while superficially attractive would
| have the undesirable effect of causing all existing TEI software to
| fail, since namespaces cannot be defaulted by e.g. XSLT processors.

I thought about this a bit on the flight home, and I think it may be
a red herring. To say that stylesheets written for P4 will not work
with P5, given that we have already agreed that backwards compat-
ibility is not a requirement, is something like saying stylesheets
written for the AAP tagset won't work with the ISO 12083 tagset --
while one may be based on the other, they are not the same.

That said, I can understand how Sebastian, who has put an enormous
amount of (much appreciated) work into very nice stylesheets for P4,
would prefer to keep it easy to get them to work with P5.