[...]
> > I set the bugreport to 'testing' and didn't yet upload a package
> > yet, please send me a short ACK if you think it's really OK. :)

Here we go:

There is one issue.
"if [[ -x $(type aterm) ] ; then"

Alexander's patch involved the old $(which aterm) solution and does
*not* include this problem. It is however in our hg repo:

Alexander's patch (which I downloaded from the bts):
[snip]
-     if [ -x $(which aterm) ] ; then
-        isutfenv && [ -n "$LANG" ] && alias aterm="LANG=${LANG/(#b)(*)[.@]*/$match[1].iso885915} luit aterm"
+     if [[ -x $(which aterm) ]] ; then
+        isutfenv && [[ -n "$LANG" ]] && alias aterm="LANG=${LANG/(#b)(*)[.@]*/$match[1].iso885915} luit aterm"
[snap]

Our repo (the changeset that incorporates the patch):
[snip]
- if [ -x $(type aterm) ] ; then
- isutfenv && [ -n "$LANG" ] && alias aterm="LANG=${LANG/(#b)(*)[.@]*/$match[1].iso885915} luit aterm"
+ if [[ -x $(type aterm) ] ; then
+ isutfenv && [[ -n "$LANG" ]] && alias aterm="LANG=${LANG/(#b)(*)[.@]*/$match[1].iso885915} luit aterm"
[snap]

Mercurial bug?
Anyway, the attached patch fixes this problem and does further
cleanups. Like == instead of =, -eq for integer comparison and doing:

[snip]
-  if [[ $foobar == baz ]]
-  then
+  if [[ $foobar == baz ]] ; then
[snap]

Note, that I am still troubled about the use of 'type' instead of
'which'. Could someone enlighten me about this?

Also, if the above issue *is* a mercurial problem, that would make me
kind of nervous, too.

Comments?

Regards, Frank