is it the future?

Dave Ihnat dihnat at dminet.com
Tue Sep 23 22:18:57 UTC 2014


On Tue, Sep 23, 2014 at 05:25:28PM -0400, Matthew Miller wrote:
> Seriously? Please review the code of conduct linked at the bottom of the
> message.

I agree that it's probably not a good response.

> This _really_ isn't the place to continually review systemd or promote
> conspiracy theories. If you want to help people work with systemd, or work
> around shortcomings (and sure, it's got plenty), awesome. If you want to
> help work on some kind of alternative system, that's awesome too.

If I may offer a suggestion--this is, in fact, exactly the place to
discuss elements of Fedora that have wide-ranging implications, such
as systemd.  Unless and until it's demonstrated that it's an optional
component, and not a required component, of the operating system, then
its acceptance in Fedora has extensive and profound implications on
functionality and operation.  Suggesting a RYO remix isn't a solution.

And, as Fedora is the testbed for Redhat and Centos, just that kind of
discussion and expression of concern, I would propose, is what Redhat
should be interested in hearing.  Without polemic, _ad hominem_
argumentative posts, or non-substantive "Is not!" "Is too!" arguments.
(Remember debate rules?).

Unfortunately, I wish this discussion had occurred long ago--long ago
enough to have had an effect on Redhat's decision, already made, to make
systemd a part of RHEL and, by extension, CENTOS.

The fundamental issues seem to be:

 o There is still a lot of concern about systemd.

 o It appears that, once systemd is incorporated, it's not an optional
   component.

 o Much of the concern is centered around the contention that the purview
   of systemd, as implemented, extends far beyond the ostensible purpose
   of revising only system initialization.

 o The maturity and stability of the systemd code base is also a matter of
   serious concern.

All of these, I would suggest, are valid concerns for the Fedora list.  The
biggest question, it seems to me, is whether it's productive to discuss any
of this at this point in time.  Redhat seems to have already decided to
throw its lot in with systemd; can any discussion here change that
decision?  If not, any effort we expend here is useless.

Let's decide that before we argue any more on the merits--or lack
thereof--of systemd itself.  If it's not going to change Redhat's
decision, then all we can meaningfully discuss here are discovered
issues and any resolutions of same with the current implementation of
systemd.

Sincerely,
--
	Dave Ihnat
	dihnat at dminet.com


More information about the users mailing list