I have a data set with Norwegian place names and words. Using character mapping the correct letter is visible like æ, ø and å. All reports in TMG work. When I export GEDCOM from TMG version 8 the characters do not map as expected. The choices I have are ANSI, ANSEL or IBMPC. The IBMPC does not support the extra characters. The other two do export "non-roman" characters but not the ones expected. I use TMG on windows 7 running under Parallels on an iMac. Have others seen this?