[issue431] wm-ng on grml-medium 0.1
Michael Prokop
bts at bts.grml.org
Sat Feb 14 21:49:05 CET 2009
Michael Prokop <mika at grml.org> added the comment:
* Frank Terbeck <bts at bts.grml.org> [20090214 20:29]:
> [snip]
> #!/bin/bash
> Seems like a plain sh script to me.
> Bash can handle that of course. But it's not needed, AFAICS.
I was afraid due to:
possible bashism in usr_bin/grml-exec-wrapper line 11 (type):
if type -a Xdialog 1>/dev/null 2>&1 && test -n "$DISPLAY" ; then
possible bashism in usr_bin/grml-exec-wrapper line 31 (type):
if type -a "$PROG" 1>/dev/null 2>&1 ; then
=> Is this really a bashism?
> # use Xdisplay only if present and running under X:
> display_info() {
> if type -a Xdialog 1>/dev/null 2>&1 && test -n "$DISPLAY" ; then
> Xdialog --title "grml-exec-wrapper" --msgbox "$1" 0 0 0
> else
> print "$1">&2
> stderr? really?
Yeah, I guess it makes sense as display_info() is invoked in error
situations only.
> # make sure to support 'grml-exec-wrapper sudo wireshark' as well:
> case $PROG in
> *sudo*) PROG="$2" ;;
> esac
> if type -a "$PROG" 1>/dev/null 2>&1 ; then
> exec $@
> Rather:
> exec "$@"
Thanks.
> else
> RC=1
> display_info "Sorry: ${PROG} not available.
> Looks like the grml flavour you are using does not ship ${PROG}. :(
> You can search for ${PROG} executing:
> apt-get update && apt-cache search ${PROG}
> "
> fi
> The whole RC variable is unneeded.
> The only way, you end up down here is if you come down from the else
> path above (which is unneeded, for the very same reason.
> Just 'exit 1' should be good enough (and much clearer).
Ah right, good catch. :)
Thanks for reviewing!
regards,
-mika-
----------
assignedto: mika
messages: 1244, 1245, 1876, 1877, 1881
nosy: ft, maddi, mika
priority: bug
status: in-progress
title: wm-ng on grml-medium 0.1
topic: release-stopper
_____________________________________
GRML issue tracker <bts at bts.grml.org>
<http://bts.grml.org/grml/issue431>
_____________________________________
More information about the Bugs-changes
mailing list