Re: question about systemd
On Sat 11 Oct 2014 at 07:38:42 -0400, The Wanderer wrote:
> On 10/10/2014 at 07:53 PM, James Ensor wrote:
>
> > On Fri, Oct 10, 2014 at 7:31 PM, John Hasler <jhasler@newsguy.com>
> > wrote:
> >
> >> James Ensor writes:
> >>
> >>> My impression is that the idea of "systemd's entanglement" has
> >>> been blown way out of proportion.
> >>
> >> The entanglement discussed here earlier had to do with the design
> >> of the Systemd suite, not with dependencies.
>
> (Well, there was some discussion about the dependencies side of thing as
> well, but I think that's more an effect of the design of the systemd
> suite rather than a primary issue of its own.)
>
> > Exactly, and that has been blown way out of proportion. You do not
> > need to have systemd installed or running to have a usable
> > Debian-testing install.
>
> Er... how do these two sentences make sense together?
>
> If I'm reading you correctly, you're claiming that the entanglement
> involved in the design of the systemd suite has been blown way out of
> proportion, and in support of that you're citing the fact that you do
> not need to have systemd installed to have a usable Debian testing system.
>
> But whether or not you have to have systemd installed to have a usable
> Debian testing system is not about the design of the systemd suite; it's
> about dependencies. So your citation seems to have nothing to do with
> the claim at hand.
>
> Is there something I'm missing that makes this make sense?
1. The design of systemd is thought to have certain consequences which
may have an inpact on a Debian system.
2. One consequence of the design is that there are 'entanglements' when
it comes to installing packages on Debian. 'Entanglements' are simply
dependencies. Debian has a package system which has relied on the
concept of dependencies for many. many years. Inventing a new term for
a familiar process doesn't appear helpful.
3. The consequence in 2. is a major consequence of the perceived design
limitation in systemd . It might actually be the only one of practical
importance to users installing and managing a Debian system.
4. It is an undisputed fact that testing and unstable users can remove
systemd absurdly easily.
5. The practical outcome of 4. and 3. is that the proposition in 1. is of
little importance to Debian users. Most are only concerned with 'Does
it work?'. 'It does indeed' is the answer. These users have little
interest in perusing the design of systemd. Even if it didn't work as
well as it does they would still have no interest.
The design aspect is trumped by reality; hammering away at it doesn't
increase its significance.
6. A number of users want a different init system. Are they catered for?
Of course they are! Please see 4.
1. was discussed earlier this year. A decision was made. Reprising it in
debian-user may produce some clarification but the fundamental framework
for Jessie has been laid down.
And to illustrate how much work Debian maintainers put in to respond to
users' concerns:
root@gnome-jessie:~# apt-get install sysvinit-core systemd-shim
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following extra packages will be installed:
cgmanager libcgmanager0 libnih-dbus1 libnih1
Suggested packages:
pm-utils
The following packages will be REMOVED:
systemd-sysv
The following NEW packages will be installed:
cgmanager libcgmanager0 libnih-dbus1 libnih1 systemd-shim sysvinit-core
0 upgraded, 6 newly installed, 1 to remove and 0 not upgraded.
Need to get 482 kB of archives.
After this operation, 1,030 kB of additional disk space will be used.
Do you want to continue? [Y/n]
What more could a Debian user want?
Reply to: