That's because logic is unambiguous, and it allows us to clarify our own thoughts. Otherwise our brain gets ridden with misconceptions, prejudices, and lousy thinking. The need of pure logic is a testament to our brains' messy nature, not any characteristic of the workings of the world.

Of course it will be useless for you if you already have some understanding of the UNIX heritage. As with all metaphors, its value is for people who know very little about the topic, in that it helps them relating the topic to something which they're already familiar with.

For someone without a previous knowledge in the history of UNIX, the metaphor provides a mental map to navigate intuitively what was perceived as an impenetrable technical mess. It can provide the idea that there is a heritage of branching from a common origin, a sense of what are the main branches, their relative antiquity and importance.

Moreover, it's funny and light-hearted. Why does everything has to have a practical purpose?

Sort of. I wouldn't call it a lie but, as it's true that all modern applications (the only parts of the computer the user interacts with) depend on the operating system to work.

s/trail/trial/
s/know how an OS /know how an OS works /

Because it conveys the right message for the people hearing it. That's why MS used it -it's a very good definition for someone who doesn't program computers for a living. MS tried to derail the trail by dumbing down the tech details, which shouldn't have been done at the Court. This doesn't mean that hiding tech details is always wrong.

End users don't get any direct benefit from the OS- it's a tool for the developers, so users don't require any detail about it's inner working; they literally don't need to know how an OS in order to accomplish their goals when using the computer.

Users only interact with the shell, the package system and maybe the file system; and those do not strictly belong to the OS, but are just applications bundled with it. And those are precisely the parts described in TFA.

That definition is not intended to win trials here, nor to be used in any technical context. Their choice of words means that their target audience is not the stereotypical Slashdot crowd.

It implies that we can install this environment to our families, and still hope to use it ourselves. Some of us believe it's a good thing, that the people who think "the beige box is the hard drive" can use computers. But not many developers know how to make a computer that they can use; making software easy to use is much more difficult than most programmers realize.

They have a main webpage with a clean design, and they explain what they do and why anyone in the target audience should care, without falling prey to corporate-speak. That alone bests more than 90% of previous desktop environments, yet is the bare minimum than any user-facing project should have. Plus, the FAQ and About pages actually explain their motivations rather than a few obscure technical details.

That "operating system, a suite of software that makes your computer run" made me shed tears of joy.

You're just describing human nature there, not an essential difference between science and religion. Most people doing science don't follow their assumptions to their ultimate philosophical consequences, and remain at a comfortable pragmatical ground.

Conversely, there are rational theologians (starting with scholasticism, which existed prior to the modern scientific method) that make all assumptions explicit - they just happen to use a different set of assumptions than positivism. This doesn't make them less rational, although it makes them less scientific.

That, sir, is the best comment in this whole thread. It's so rare to find a real True Neutral alignment this days...

Exactly. And most of the time, disproving their whole axiomatic framework through logic is simply impossible to do, as that framework is self-consistent.

Note that this is true for religious beliefs, but it's also true for the scientific method and rationality - their core assumptions are non-falsiable.

I change my belief. This is what a rational being does

In that case, that "rational being" thing must be a mythical creature. Most of the time, human beings create rational explanations that match their pre-existing beliefs and subsconscious decisions they've taken, based on the emotions they evoke; not the other way around.

The shorter way to say that is rationalization, and it has a biological basis that has been studied through magnetic resonance imaging. There are some beliefs that can be discredited by careful assessment of axiomatic frameworks, looking for inconsistencies in them, but certainly the scientific method does not apply to non-falsifiable ideas like core religious beliefs; if you apply logic to them, you only get more and more complex and convoluted scholastic theories.

If your search engine doesn't apply "smart", in what order do you suggest it to return its results?

Free Google hint: try the verbatim search option - you get the smart ordering minus the smart word guesses.

Works -> world. Damned auto-correction.

+1 to that.

Even when we can improve our tech and evolve our tools, there's something to be said about taking the slow route. In nature, evolution creates generalist species that can survive to works cataclysms, when specialised species that were "better" for the old environment perish en masse.

I think we know how to build such better tools for specific uses, but we have no clue about how to create "generalist", all-purpose resilient tools. Understanding those may very well require a major breakthrough or two in our knowledge of evolution. Its likely that we'll be able to build them before we know how we did it.

we already know the most obvious tricks it developed at the lowest level, and almost nothing at the intermediate levels

There, fixed that for you

the difference is that we can improve the process itself by applying recursion to it, which is why all our technologies go through a period of exponential improvement, while evolution's process remains the same old linear technique and doesn't change

Fair enough to that too. It's clear that evolution has worked so far through brute force and random environment changes, and humans can generate directed evolutionary environments to accelerate the process. Although humans applying their wetware to solve specific problems can be seen as nature itself applying a "recursive ability so as to improve its own methods", from a certain point of view. ;-)