[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Legitimate exercise of our constitutional decision-making processes [Was, Re: Tentative summary of the amendments]


Florian Lohoff:
> There are tons of people who think that all the above functionality does
> not belong to a init systemd or ecosystem.
There are also tons of people who could care less, as long as it gets the
job done.

Then there are tons of people who are _very_ happy about the fact that
systemd enables them to do a significant part of their job a whole lot
faster and safer.

There are also tons of people who are simply lazy and think that every
single line of code that the existence of systemd and its ecosystem allows
them to remove, or not write in the first place, is a net win.

Arguments by popularity are not going to sway anybody here. Otherwise I
could shut you up with a simple "most other distros have switched and are
mostly-happy with it". :-P

> It will push
> our users, lobbyists and con-systemd devs away.
Good. The rest of us can then finally focus on our real job(s) again.

> The problem starts with naming all of them systemd-foobar.

You know what? (a) not all of them are, (b) you avoid naming conflicts that
way. Tools which are designed to be universally installable simply can't
have name collisions. I'd very much like to use "nspawn" or some other
cool short name in place of "systemd-nspawn" every time I start a process
in a chroot, but the idea that somebody is already using that command name
for something else isn't too far-fetched.

Thus, IMHO this is a feature.

Of course, if you _insist_ on treating this like an intentional ego-boost,
instigated by systemd's creators, then I can't stop you, just as you can't
prevent me from thinking that you're silly.

> A lot of the oldtimers (like me) would agree to this which i read in rant about systemd:
Arguments by ranting are not going to convince anybody either.
>    Because confident young men in a hurry to make their own mark on the
>    world have little time for learning the tools or the lessons of the past."
If Lennart's string of talks about systemd at various Linux conferences
during the last couple of years had been a one-way street, he'd not have
been re-invited in the first place, and nobody would have adopted the

Also, some of us somewhat-old-timers think that the lessons of the past
(and the desire to finally do better, after futzing around the problem
space's periphery for the last twenty years or so) are precisely the reason
why they *like* (most of) what systemd is doing.

> Every time someone tells me about the shiny new features i must think of this.
Every time I read a rant like this, I think that the ranters are unable to
conceive of the fact that people (in particular, Lennart) are actually able
to learn from their mistakes.

Systemd is _not_ like pulseaudio in its infancy.

-- Matthias Urlichs

Reply to: