I put ConTeXt on a new server. It works fine except that it is not aggregating index results. I have isolated it to a single line, but i do not understand why it breaks indexing. Feedback from anybody would be immensely helpful. I have already spent way too much time figuring this out. %Minimal example: % ConTeXt ver: 2017.05.15 21:48 MKII fmt: 2017.8.17 int: english/english \enableregime[utf] \starttext \index{Aardvark}Aardvark \index{Baboon}Baboon \index{Camel}Camel \pagebreak \index{Baboon}Baboon \placeindex[compress=yes] \stoptext Outputs Index like this: A Aardvark 1 B Baboon 1 C Camel 1 B Baboon 2 and issues warning when compiling: references : duplicate reference [][Index:b] on page 2 However, if i remove this line, it works fine: \enableregime[utf] I need that line to be able to handle diacritics in the text however. I cannot upgrade to mkIV yet as there is a complex toolchain involved. Running this example with the same installation files and same input returns different results when run on my local workstation versus this new machine i am working on. It also works fine on the live.contextgarden.net tool. Can anyone think of a reason why adding "\enableregime[utf]" would break the index aggregator? File permissions? Missing tool? Something wrong with fonts? Some other context-sensitive misconfiguration? Any help would be appreciated. Thanks a bunch.