Mailinglist Archive: opensuse (982 mails)

< Previous Next >
Re: [opensuse] Getting rid of systemd and putting sysv back
Andrei Borzenkov wrote:
В Sun, 28 Sep 2014 13:21:54 +0200
lynn <lynn@xxxxxxxxxxxx> пишет:

On 28/09/14 02:01, Dirk Gently wrote:
Anton Aylward wrote:
On 09/25/2014 11:49 AM, ianseeks wrote:
On Wednesday 24 Sep 2014 15:43:55 Dirk Gently wrote:
Anton Aylward wrote:
On 09/24/2014 02:15 AM, Felix Miata wrote:
I don't think the same
about systemd so much any more after reading a few days ago this:

That's excellent.

Those 17 points sum up 'the UNIX way' very well, much better than the
rants we've seen here, which have mostly been recidivist.

The claims that systemd is undocumented and similar are quite

If it's not in the man pages, it's not documented.

man systemd - works for me

Not only that!

$ apropos systemd| wc -l

aaron-as-dirk's complaint about getting the GUI to work is frivolous.
The post-systemd problems I've had with Xorg have been the same ones I

And if SystemD FAILS before your GUI comes up (i.e. unrelated to the GUI),
then what, Mr. Not-as-clever-as-you-think-you-are?

OK. I'll take my opportunity whilst it lasts.

It does fail e.g. /run/user/0 is not created for the Kerberos root cache
so nothing is mounted from AD. The problem is, no one at either SUSE nor
openSUSE understand the problem. The bugs never get fixed (or read even)

There were at least two suggestions how to workaround it; and you
yourself have confirmed that at least one of them works. So it is not
that a) nobody reads them and b) it is something that is broken beyond

The real underlying problem is that having filesystem as common cache
for unrelated process is fundamentally incompatible with idea of name
spaces. The only fault of systemd here is that it makes use of name
spaces more streamlined. But there is no way to ensure that /tmp/krb5cc
created by one process will be accessible by another process.

Or in OTHER WORDS, SystemD is using a model which is fundamentally
incompatible with the entire rest of Linux.

Hmmmm... So... we should change EVERYTHING because Poettering and
Sievert want Microsoft Linux ????
What's next... a fucking slow-ass registry, too?

because no one either at SUSE nor openSUSE has the resources or
understanding to fix them. This is a pure systemd error which has
knocked out working with SLES or openSUSE after systemd in a domain
completely. The only solution is a workaround. 844198 has been open for
almost a year. No one has addressed the issue.

Major changes in 1.12 (2013-12-10)

Add collection support to the KEYRING credential cache type on Linux, and add
support for persistent user keyrings and larger credentials on systems which
support them.

Factory (13.2) has 1.12.2. Did you look whether problem still exists
there? If yes, did you test whether using kernel keyring solves it?

And let's face it, if
Fedora have problems with the same issue, what chance do we stand?
L x

To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
To contact the owner, e-mail: opensuse+owner@xxxxxxxxxxxx

< Previous Next >
This Thread
Follow Ups