> We *could* of course just invent random Node names, but that has the > downside that we would then need to either use XPath queries to > locate page, or maintain an internal normalized name -> uuid > mapping, and that's extra baggage, I think.
I think so, too. In addition, AFAIR XPath queries have no future in JCR. JSR-283 only contains them for backwards compatibility reasons. Another point which is just coming to my mind: What about cutting out double blanks? Regards Florian
smime.p7s
Description: S/MIME Cryptographic Signature
