cvs --> mercurial: Configuration error: 'ascii' codec can't encode character u'\xdc'

[ThurnerRupert]

what could make tailor think reverting to "ascii" is wise?

17:19:09 [C] Upstream change application failed
Configuration error: 'ascii' codec can't encode character u'\xdc' in position 11: ordinal not in range(128): it seems that the encoding used by either the source ("ISO8859-1") or the target ("UTF-8") repository cannot properly represent at least one of the characters in the upstream changelog. You need to use a wider character set, using "encoding" option, or even "encoding-errors-policy".

we also tried to hardcode the encoding to UTF-8. the effect was that source and target are stated as "UTF-8" (which is not true) and the same "ascii" error.

the log message of that change was empty, the only strange thing i see was that a windows cvs client removed a file called "~$$CSS Übersicht.~vsd"

cvs to mercurial synch happens on unix, creating the file with escaping by "\" works as shown below:

$ locale
LANG=
LC_CTYPE="en_US.ISO8859-1"
LC_NUMERIC="en_US.ISO8859-1"
LC_TIME="en_US.ISO8859-1"
LC_COLLATE="en_US.ISO8859-1"
LC_MONETARY="en_US.ISO8859-1"
LC_MESSAGES="en_US.ISO8859-1"
LC_ALL=en_US.ISO8859-1
$ touch "~\$\$CSS Übersicht.~vsd"
$ ls
~$$CSS Übersicht.~vsd        

hg version is 0.9.1, or 0.9.2 (with patch from #86).

[edgar.alves]

I've already sent the patch to lele, but here is the diff in case you're in a hurry:

vcpx/changes.py

@@ -54 +54 @@
         else:
             s += ' from ' + self.old_name
         s += ')'
+        if isinstance(s, unicode):
+            s = s.encode('ascii', 'replace')
         return s

[lele]

I applied Edgar's patch, as in [1303].

[ThurnerRupert]

thanks a lot, it works now!