Originally Posted by
colouredmirrorball
Now here comes the embarassing part. My program writes this extra header, but instead of three zeros, it writes the ascii character '0' which is equivalent to the number 48. This is a really silly mistake I should have caught much earlier, but my program could read the faulty files just fine by some coincidence... Of course, a lot of software will have trouble with reading the resulting files.
Anyway, this bug is fixed and I plan on releasing a new version soon. Files produced with the faulty version will be able to get loaded in again but will contain some extra points as the invalid header will be considered as regular data. There are some unique new features which don't really work yet, so I kind of don't want to release it now. But if somebody needs an emergency fix, I'll hapily provide a developer version.