Community
Participate
Working Groups
I frequently compare local sourcefiles with the latest versions from our CVS repository, mostly by double clikcing a conflict in the synchronize perspective. The compare editor then loads the local file using the specified character encoding for this file (in my case latin1) and the remote file using the default character encoding of the workbench (in my case utf -8). The result is a scrambled remote file and a lot of false positive changes. The remote file should always be read in the same character encoding as the local file. Setting the default character encoding of the workbench to latin1 has helped for my project, but this was not obvious and took me some weeks to discover. It is also a problem if two different encodings are used frequently.
*** This bug has been marked as a duplicate of 72995 ***