Re: [gentoo-user] Debian just voted in systemd for default init system in jessie

["Yuri K. Shatroff" <yks-uno@yandex.ru>]

24.02.2014 02:32, Alan McKinnon wrote:
> On 23/02/2014 20:18, Canek Peláez Valdés wrote:
>> I don't think forking would attract much developers. Writing something
>> new trying to follow "the*nix design principles", but being modern and
>> with the same features (all of them optional, of course) of systemd
>> will have more chances; although I think it will fail because most of
>> the people that can code "better" actually like the systemd design,
>> and would prefer to contribute to it.
>>
>> And if you found enough of this mythical good-coders, good luck
>> defining what it means "the*nix design principles".
>
>
> I've been wondering about this concept of "the*nix design principles"...
>
> I've now concluded it's a myth, much like invisible pink unicorns.

I may not be an authority, too. But please allow me to refute your 
arguments.

> Is it like the kernel? A huge monolithic chunk of code with support for
> modules?

It ain't. No monolithic chunk of code, it's configurable.

> Is it like X11? A huge monolithic chunk of code that has a bizarre build
> system for years, and took something like 5 years of hard work to get it
> modular? And is 20 years behind the times? And *still* requires devs to
> jump through hoops to get a rendered image through a compositor and back
> up the the GPU?

It's grown to that, but in the beginning it was (striving to be) a clean 
system doing generally one thing (graphical client/server) and doing it 
well.
[1]
It's not X11 devs' fault that GPUs and all that multidisplay/ multimedia 
stuff don't work well with client/server arch because they were designed 
for some other, you know which, OS. I assume if the GPU vendors had 
their specs opened "20 years" ago, some wayland-like stuff would have 
been ready near that time.

> Is it like perl? Support every possible way to do something if it
> remotely makes sense to do it, no matter how bizarre the syntax?

Perl (I suppose you know what it stands for) is great (probably the 
greatest) for what it was invented for: text manipulation/analysis. It 
could have been a good replacement for many things like awk, sed, tr 
etc. if the author were less ambitious to conquer the world with Perl.

> Is it like python? Pick ONE way to do it and stick with it dammit!

You misquoted. The phrase is: "there should be one—and preferably only 
one—obvious way to do it", *one* meaning 'at least one', complemented 
with *should be* and *obvious*.

> Is it like php? Do whatever you feel like?

Php was a Unix design? LOL. Php wasn't a design at all. It was just 
another personal home pages perl script.

> Is it like command line text processing tools that only do one narrow
> thing well? [1]

Perfectly well.

> Is it like bash? I can't find a decent description of how bash came to
> be except it's like Vogons - wasn't designed and didn't evolve, it just
> sort of ... congealed

Bash or sh? What about ksh, csh, zsh etc? Well, a shell actually does 
two things: interactive shell and scripting. Let's ponder on how they 
can be separated?

> Not to rain on anyone's parade, but there's a prize of 40 internets up
> for the first person who can clearly and unambiguously define "Unix
> design principles" with specificity so that it is globally applicable.

A truism: There's nothing globally applicable.

> Best I can come up with is "Use common sense and build stuff that can be
> used and maintained" which is wonderfully descriptive but really sucks
> as a definition.

Something like this, but neither is it globally applicable.

>
>
> [1] For lack of a better term, let's just call systemd here a "system
> controller". What is this ONE thing a system controller should do and do
> it well?

An init daemon generally does one thing well.


[1] http://en.wikipedia.org/wiki/X11#Principles

-- 
Regards,
Yuri K. Shatroff