JEP 400 changes the result of Charset.defaultCharset()
from the system's default encoding to UTF-8.
This has no effect on most Linux systems,
and I think it is fine for CPAchecker if its textual output files
are always in UTF-8.
However, this also changes what encoding is used when communicating
with other processes via stdin/stdout
(on non-UTF-8 systems like Windows, at least)
and this might be a problem, because other tools likely expect the
system's default encoding to be used for stdin/stdout.
Of course, the problem is likely to be small for CPAchecker
because non-ASCII characters are expected to occur rarely,
but it is still better to be safe.

This change mainly updates SoSy-Lab Common to a version
where ProcessExecutor always uses the system's default encoding,
even on Java 18.
Furthermore, we add a warning if users manually fiddle with encoding
settings, because this would mean that output files would no longer be
consistently in ...