[26705] in Source-Commits
Re: /svn/athena r25695 - trunk/debathena/debathena/locker-menu/debian
daemon@ATHENA.MIT.EDU (Jonathan Reed)
Sat Aug 4 19:52:52 2012
Mime-Version: 1.0 (Apple Message framework v1084)
Content-Type: text/plain; charset=us-ascii
From: Jonathan Reed <jdreed@MIT.EDU>
In-Reply-To: <alpine.GSO.1.10.1208041939410.22690@multics.mit.edu>
Date: Sat, 4 Aug 2012 19:52:48 -0400
Cc: source-commits@MIT.EDU
Message-Id: <155D7538-2995-40FE-BECC-7B42B70337C6@MIT.EDU>
To: Benjamin Kaduk <kaduk@MIT.EDU>
Content-Transfer-Encoding: 8bit
>> if [ "$RET" = "true" ]; then
>> echo "Using real icons for debathena-locker-menu..." >&2
>> for i in $ICONS; do
>> - if ! cp -f "$ICON_NONFREE_DIR/$i" "$SHAREDIR/$i"; then
>> - echo "Can't find real icon for $i, falling back to generic" >&2
>> + if [ -f "$ICON_NONFREE_DIR/$i" ] && \
>> + ln -nsf "$ICON_NONFREE_DIR/$i" "$SHAREDIR/$i"; then
>
> Isn't the logic on the check for ln's return value inverted?
Indeed it is.
>> +case "$1" in
>> + purge|remove)
>> + for i in $ICONS; do
>> + if [ -L "$SHAREDIR/$i" ] &&
>> + ! [ -f "$(readlink "$SHAREDIR/$i")" ]; then
>> + rm -f "$SHAREDIR/$i"
>> + fi
>> + done
>
> I had in mind just unconditionally removing the link, but this is fine, too.
This felt more in compliance with policy, but I guess we can just scribble over it, since it's in /usr/share anyway. Although, hrm, if I do this in the postrm, I would need to remove the directory too, right? Otherwise it would still be lying around, since it would not have been removed in the cleanup as it was nonempty?
-Jon
