:: Re: [Dng] Which source version fo s…
Top Pagina
Delete this message
Reply to this message
Auteur: James Powell
Datum:  
Aan: Jaret Cantu, dng
Onderwerp: Re: [Dng] Which source version fo systemd are you stripping code from?
It's a bad thing when no talent idiots refuse to learn even the simplest of standards and do things against the UNIX way, the sane way, the simple way, and nobody cares to question them except the few.

Sent from my Windows Phone
________________________________
From: Jaret Cantu<mailto:jaret@realitysend.com>
Sent: ‎5/‎7/‎2015 7:46 PM
To: dng@???<mailto:dng@lists.dyne.org>
Subject: Re: [Dng] Which source version fo systemd are you stripping code from?

On 05/07/2015 09:59 PM, James Powell wrote:
> Etcnet was very involved with scripts which was always a problem for
> Red Hat. Many Red Hat scripts have been known for being substandard in
> quality and reliability, possibly a reason why they wanted systemd so bad.
>
> However, any daemon can work well if proper scripting is applied and
> mistakes are corrected properly.

Scripts are just the easiest thing to blame since everyone can see them
and far more users are capable of reading (and thus critiquing) them.

Scripts are no more different than any other piece of code that makes up
a system, but they are a lot easier to modify and understand (and thus
critique -- wait, I already did that one) to suit to your individual
purpose. That is why scripts have been strewn across *NIX since the
dawn of time. (Or the Epoch. Close enough. That's the one that
matters, anyway.)
Scripts are a fundamental building block across many aspects of *NIX, so
to focus on just init is a bit tunnel-visioned.

If a script is bad, send a patch. And getting rid of scripts altogether
won't fix any problems of quality or reliability; it will just put any
badness in the binaries where you can't easily
understand/modify(/critique!) any issues that an init system
who-will-remain-nameless might (read: will and has) caused.


I for one like scripts: start, stop, _etc_. Same scripts work for
SysVinit or BusyBox or conceptually any arbitrary init system that will
have them. I mean, the convention is part of the Linux Standard Base,
fer cryin' out loud!

http://refspecs.linuxfoundation.org/LSB_3.1.0/LSB-Core-generic/LSB-Core-generic/tocsysinit.html

Using some strange new non-script init file format is typically what
sours me on many of the alternative init systems. Pretty much every
service already has an initscript for itself available, and if not, they
are very easy to make. You'd think someone would focus on
fixing/streamlining/refactoring the oft-called crufty init executable
instead of struggling with an file format when one already exists and is
widely available for use. Simple, yes, but when has KISS been a bad
thing before?


~jaret
_______________________________________________
Dng mailing list
Dng@???
https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng