В Wed, 17 Sep 2014 18:29:35 +0200 Pavel Raiskup <prais...@redhat.com> пишет:
> On Wednesday 17 of September 2014 20:06:50 Andrei Borzenkov wrote: > > В Wed, 17 Sep 2014 13:40:11 +0200 > > Pavel Raiskup <prais...@redhat.com> пишет: > > > > > On Wednesday 17 of September 2014 11:50:08 Zbigniew Jędrzejewski-Szmek > > > wrote: > > > > On Wed, Sep 17, 2014 at 08:52:00AM +0200, Pavel Raiskup wrote: > > > > > Thanks, I agree, and yet another rule on top of that -- the /usr/lib > > > > > dropins _are ignored_ if /etc service overriding the one in /usr/lib > > > > > exists [1] which seems like it is expected (and I appreciate that > > > > > behavior). > > > > > > > > Sorry, but the .include mechanism is deprecated (and almost unused > > > > nowadays) > > > > > > Note that the thing that .include changed its semantics due to drop-in > > > feature is not really proper deprecation. Nor any documentation, warnings > > > about its usage, etc. > > > > I reread your original post and I still fail to see what has > > changed. IIRC ".include file" always meant "literally inline content of file > > in place of .include line". This is exactly what still happens now. Do > > you have pointers to old documentation that stated something different? > > Say that A.service .includes B.service. Because B.service was correctly > inlined before, A.service was able to re-define results of .include. > *file* B.service is still correctly inlined. > Pavel > _______________________________________________ systemd-devel mailing list systemd-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/systemd-devel