Re: What is ‘OS’? (was: Why "GNU/Linux" is not accepted: an observation)

[Dmitry Alexandrov]

Akira Urushibata <afu@wta.att.ne.jp> wrote:
> Subscribers of this mailing list know what an operating system is.

Yes, they for sure know.  But I would not sure, that they _agree_ on what ‘OS’ 
is.

If we exclude marginal ones (like OS == kernel), I am aware of two consistent 
definitions of ‘operating system’:

1. Operating system is a set of auxiliary programs, necessary to use 
applications, which are the programs that you actually want to use to solve 
some tasks.  Where these programs come from is irrelevant.

Since the criterion is purely subjective, there is no precise boundary: if you 
perform some scientific calculations and also use, say, youtube-dl(1), the same 
Python can be both application and a part of the OS for you.

2. Operating system is set of programs, installable on the top of hardware, 
distributed by a single vendor.  In other words, it is exactly the same as 
distribution (‘distro’).

Despite that this criterion is objective, the same program still can be both: 
Minesweeper™ and KMines are part of operating system as long as they are 
distributed within MS Windows and Debian respectively, but are not if installed 
separately.

These two definitions are not unrelated, though.  You might easily come to idea 
to install the latest Python from upstream if you write in it, and hardly 
bother to get to know, in which language youtube-dl is written, at the same 
time.

So feel free to mix them in right proportions to construct a desired discourse.

Back to the question of (dis)agreement, though.

Those, who like to talk about GNU (on Linux or otherwise) operating system as a 
 fait accompli, are evidently inclined to the first one.

While those who were going to dub Guix, the distribution, ‘the GNU OS’ few 
years ago in order to finally fulfil the GNU Manifesto, seem to prefer the 
second.

signature.asc
Description: PGP signature