Some filetypes, such as text files, do not identify their character encoding. When a file of this type is attached to Confluence, Confluence must assume it uses the default character encoding specified by the JVM.

If one of these files has a different encoding from the default, the following problems occur:

1) Searches for the contents of the file may not return a result, as any characters not matching the JVM-specified encoding will not be indexed
2) When the file is viewed in the browser, it will be shown in the JVM-specified encoding. This will garble any characters not matching the JVM-specified encoding
3) When the file is downloaded, the user's system will assume it is in that system's default character encoding. If the encoding differs from the original encoding, this will garble any characters that do not match