Wednesday, May 1, 2013

Systemd: An Accident Waiting to Happen

by Dietrich Schmitz

Over the past week or so, I have been following events surrounding the ongoing work being done on systemd by Red Hat employee +Lennart Poettering.

Personally, I have been trying to figure out where the safest place to be in the Distro space would be after Fuduntu closed its doors.

Looking at the situation and taking into consideration various viewpoints, it would seem that the criticisms being waged against systemd adoption are not unfounded.

In a recent comment thread on one of my stories Red Hat is Here Today and Now, Fuduntu Founder Andrew Wyatt zeroed in on what he deems is the central issue surrounding systemd with this comment:

"If the work on systemd doesn't stand on its own merits then work will amplify to simply avoid it." 
Impossible, remember they merged it into udev making it a requirement. They were able to sidestep the technology being adopted on merit by merging it. This allowed them to force it onto every system be it sysvinit, upstart, openrc, or whatever because they all used udev - it was the common denominator. 
"Yes, udev on non-systemd systems is in our eyes a dead end, in case you haven't noticed it yet. I am looking forward to the day when we can drop that support entirely." - Lennart Poettering ( 
If (insert big bad corporation here) had done such a thing, it would have been considered monopolistic behavior. RedHat seems to somehow get a pass however for its predatory behavior. 
"I am not convinced that will happen in as the level of skill required to solve such problems is not within reach of many developers who are simply avoiding low-level programming and following the pack." 
System initialization and process management are not really complex problems. The problem had been solved with simple shell scripts for what, 30 years?

I think this is reasonable--quite reasonable thinking in fact.

It worries me that Red Hat has sponsored such work that flies in the face of common sense concerns.

Today, I opened up shop with my first cup of coffee to find a comment in my private Linux Advocate community from +Chris Ahlstrom giving me a link to a post made by Leszek Urbanski back in 2011.  It is I feel perhaps the most credible critique I have read thus far concerning systemd.

Reading Urbanski's criticism leads me to believe that somehow one Man's design objectives have been allowed to take Linux and hold it hostage to his whimsical control.

How is this possible?  It concerns me greatly that there was what appears to be a total lack of concensus building around even attempting to have agreement that systemd was worth pursuing before getting the project underway.

The result now is that systemd has taken hold and it has rooted itself firmly in place with at least a half dozen major Distros.

I regret that systemd has been allowed to muck around with things that have worked for years with great flexibility.

Read the Urbanski post.  I think most intelligent people will agree with his logic.

Is there still time to make a change?  I hope the Linux Community will come to their collective senses and see systemd for what it truly is.

Systemd in my view is 'an accident waiting to happen'.

-- Dietrich

Enhanced by Zemanta


Post a Comment