The README of nfs-utils explains that we must not notify clients
before nfsd is running, otherwise they may fail to reclaim their
locks. OTOH it's allowed but not required to start "rpc.statd
--no-notify" before nfsd. So for simplicity we do both after starting
nfsd.
Turns out that remote-fs-pre.target is not actually "wanted" anywhere,
so statd is not started before remote filesystems are mounted. But
remote filesystems do "want" network-online.target, so we can use that
to pull in statd and idmapd.
Not sure if this is really the right thing to do, but it works for
now. Background:
https://bugzilla.redhat.com/show_bug.cgi?id=787314http://hydra.nixos.org/build/5542230
The changelog is available at:
http://www.cmake.org/files/v2.8/CMakeChangeLog-2.8.11
I had to create a new patch from scratch, that's why it contains a lot
of differences. CMake developers has rewritten the files
(Modules/Platform/Linux.cmake and Modules/Platform/UnixPaths.cmake) to
comply to their coding style requirements and a lot of elements has
switched from upper to lower case.
Also, the previous patch partially consisted of commenting some
instructions mixed with line removals whereas mine is directlty
removing them in order to avoid useless evaluations and parsing in
resulting files.
When nixos-rebuild grabs a new kernel, it will build new spl/zfs
modules, which will change the service. On completion nixos will try and
restart the services which will try and import pools again, and
generally will fail.
The pools are already imported, we don't need to do it again..
I also added a patch that makes dovecot search for
plugins in /var/lib/dovecot/modules. This way, you
can add plugins from several packages without running
into circular dependencies. The module dir needs to
be populated before the dovecot service is started,
for example. This is currently not done in NixOS, so
you need to implement your own service in order to
get the plugins working.
The module patch has not been added to the old 2.1.x
package.
If no config.pulseaudio is explicitely set to false, build with pulse
support, because even if there is no pulse server available, chromium
will fall back to using ALSA.
And we definitely want to avoid that users have to build chromium for
themselves just for the sake of having pulse support. Thanks to @devhell
for actually helping me discovering this (I for myself do always rebuild
Chromium, so I won't notice those kind of things).
Signed-off-by: aszlig <aszlig@redmoonstudios.org>
The commit downgraded the german language dictionary from version 20030222-1 to
20120607 (and broke the build in the process). That change was probably not
intended?
Eelco Dolstra