As Henrik Ibsen's Writings (HIW) are now moving from the published 
printed book to the electronic publication we are using this as a chance 
to tidy up some of theb old encoding. HIW started out in SGML as TEI P3 
documents, this was changed and all the files updated to P4 XML a little 
more than 10 years ago, and now we hope to move it all to P5.

The documentation here:
http://www.tei-c.org/Guidelines/P5/migrate.xml
is handy, and our plan is to change the way we have modified TEI 
elements: now our idea is to place them in our own namespace so that 
everything that claims to belong to the TEI namspace is unmodified TEI. 
This requires some work, but apart from that, I wonder if tyhere have 
been some subtle changes between P4 and P5 that are not listed in the 
document pointed to above. So far I seem to see a difference in the 
definition of respStmt (and possibly bibl, but I have to look further here):

According to the TEI P4 documentation at the TEI website, respStmt is 
defined like this:

<!ELEMENT respStmt %om.RO; (resp | name | %m.Incl;)+ >
<!ATTLIST respStmt
       %a.global;>

Now this seems to say that the following is legal in P4:

<respStmt>
   <resp>some resp</resp>
   <name>One person</name>
   <name>Another person</name>
   <resp>another resp</resp>
   <name>Another person</name>
</respStmt>

This is not acceptable in P5 where the content definition is:

( ( resp+, model.nameLike.agent+ ) | ( model.nameLike.agent+, resp+ ) )

I don't know if I am now walking along an old discussion, and I have no 
problems with the new structure defintion, but this might then be 
mentioned in the document on P4 -> P5 migration.

Best regards
Espen Ore
University of Oslo