As I now
noticed a problem with the necessity of user input (for
interrupt the
screen saver) you reported also in the bug report 1615. It is
possible that
bug fixed in commit cbbc7ad0 could be related also with NFS on
/home?
Can you try it?
To test, I believe we have to rebuild tdebase with commit cbbc7ad0
and rebuild tdelibs with commits 9d76cb9 reversed and possibly
ba6b2079 reversed.
That is, to test whether commit cbbc7ad0 resolves the NFS problem
as well, we need to restore the ability to autosuspend in
tdepowersave when $HOME is an NFS mount. Currently that ability is
disabled in tdepowersave through commit 9d76cb9.
What do you think?
If, after testing NFS $HOME with commit cbbc7ad0, bug report 1623
still exists to address the NFS mount problem. The proposed
solution in 1623 is to study the KDE powerdevil code, which uses
upower.
I tried for a few hours trying to rebuild packages with reversed
patches, in the hopes of restoring autosuspend when $HOME is a
network mount. I gave up as there are too many changes for my C++
skills to overcome.
We can't test whether commit cbbc7ad0 resolves the problem until
the autosuspend feature is restored to tdepowersave when $HOME is a
network mount.
I really want to see autosuspend work with tdepowersave when $HOME
is a network mount. If you can restore autosuspend when $HOME is a
network mount, then I'd be happy to retest tdepowersave with commit
cbbc7ad0.
Darrell