Re: Choose your side on the Linux divide
(Sorry, Lisi. This goes beyond losing or winning gracefully.)
On Fri, Aug 29, 2014 at 8:21 AM, AW <debian.list.tracker@1024bits.com> wrote:
> On Fri, 29 Aug 2014 07:45:12 +0900
> Joel Rees <joel.rees@gmail.com> wrote:
>
> > Are you a sockpuppet arguing ironically? If so, which side do you
> > support? Microsoft?
>
> I guess you're not interested in the previous post about ad-hominem...
Ad-hominem, in logic, is when you attack the bearer of the message
instead of the message.
I think you understand that, but choose to ignore it because you would
have to examine your logic.
> and... Oh! You mean unix sockets... hahaha! that's funny.
Cute.
About as cute as erasing the evidence so you can complain that someone
insults you by questioning both your logic and your purpose.
If your purpose were clean, I think you would find less need to defend
bad logic.
> I'm on the FOSS side.
Fat chance I'm going to believe that after this thread.
> However, systemd is much better than sysvinit.
Says a lot of people who think that thinking it is will make it so, as
opposed to actually learning what the undergirdings that made *nix a
strong system in spite of apparent technical disadvantages from the
word go.
Speaking of thinking, have you read The Music Man? Do you understand
what the "think system" was a metaphor of, in the context of the times
when it was written? Do you understand that the blog you quoted from
basically ignored the bulk of criticism and analysis on the play,
including the author's own analysis?
> And
> change has been in the works for the better part of the last 15 years -
Change is always in the works. That's a null argument, completely
irrelevant to systemd's virtues and detriments.
> that's
> half of the purported 30 years of 'stable and unwanted change' of sysvinit. Look
> up simpleinit for what looks like a decent precursor to systemd.
What on earth is that supposed to mean?
If you intend to propose that the current systemd is derived from
something that was doodled on fifteen years ago, the basics of the
design of systemd goes way back before then, before the births of the
coders that are the principle authors of the current systemd. It was a
bad idea then, and the fact that CPUs have all this extra power to
waste cycles doing that kind of stuff now does not make it good design
now.
>Furthermore
> there are Debian whitepapers published around circa 2002 regarding the short
> comings of sysvinit and the advantages of switching it out for a dependency
> based init.
There are many ways to implement dependency checks in an init system.
They all have the failing that the init system has to understand what
each and every daemon must do if some need is not fulfilled. It takes
arrogance way beyond hubris to believe that one could develop a
database to track all the different needs and options of various
useful and important daemons. systemd goes beyond offering a database.
To put that in plain English, each daemon has to decide for itself how
to respond when something it depends on is not available, especially
when waiting for a dependency results in deadlock. (Theodore Ts'o
pointed that out in the discussion on simpleinit, if you look that
up.)
But Lennart Poettering is arrogant enough to think he can figure that
out and encode it deep in the internals of systemd. At absolute best
interpretation, he's proclaiming that systemd can absorb the knowledge
necessary in a timely manner, I suppose by bug reports and really fast
response time. (You think?) In a less rosy interpretation, he proposes
a time beyond which he will arbitrarily declare any new daemon that
does not follow some established pattern of initialization unworthy to
exist until he can get around to fixing his baby to know what the new
daemon needs.
> > Follow-ups, if any, on off-topic, please.
>
> You seriously must be joking if you think you'll just drop a personal insult on
> me and then sweep me off to off-topic...
Hmm.
Should I go back and see how many times you have tried to "sweep" this
thread off-topic? Or are you not one of those who have been calling
for an end to the conversation on debian-user?
> --Andrew
>
> References:
> https://www.kernel.org/doc/ols/2002/ols2002-pages-176-182.pdf
>
> http://www.safe-mbox.com/~rgooch/linux/boot-scripts/
>
> https://people.debian.org/~hmh/debconf2/debconf2-initscripts-bkg.pdf
https://lists.debian.org/debian-devel/2003/04/msg02050.html
--
Joel Rees
Be careful where you see conspiracy.
Look first in your own heart,
and ask yourself if you are not your own worst enemy.
Reply to: