Latest systemd news

[Paul W. Frields]

On Tue, Dec 02, 2014 at 05:34:53PM +0100, Lars E. Pettersson wrote:
> On 12/02/14 15:47, Paul W. Frields wrote:
> >>Well, to be frank, it was sold to solve the issue of start up time of
> >>computers.
> ...
> >This is incorrect.  The actual problems systemd was meant to solve
> >include but are not limited to:
> 
> No, it was sold to us, in the beginning, with the promise of faster boot
> time. I remember the often long and winding discussions on whether this
> materialized or not...

I think "promise" is not correct here either.  "Potential" perhaps.
As Lennart's original blog shows, the driving requirements were better
parallelization and on-demand startup, responsiveness to dynamic
system change, socket activation, better process tracking, execution
environment control, and improved event logic.  Speed of boot time was
not guaranteed but a likely outcome for some hardware configurations
(not all, which would be silly to promise, and impossible to provide).

You can read it for yourself here:

http://0pointer.de/blog/projects/systemd.html

The blog of course mentions speed, because that's what many people
were discussing at the time, but it correctly places this in the
context of what is desirable, rather than what is required.  An init
system with the proper requirements would likely yield the speed some
people were looking for, but the purpose of systemd is not simply to
be fast.

-- 
Paul W. Frields                                http://paul.frields.org/
  gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233  5906 ACDB C937 BD11 3717
  http://redhat.com/   -  -  -  -   http://pfrields.fedorapeople.org/
    The open source story continues to grow: http://opensource.com