Some general principles of naming

[Richard Stallman]

You may hear people say that an operating system is normally named
after its kernel, and therefore the "Linux" operating system
should be named after its kernel.

Actually operating systems are just about never named
after their kernels.  It is normally the other way around.

You may hear people argue that the name GNU/Linux should not be used
because the name GNU/Linux/X/BSD/TeX/X/Lynx would be impractical.  On
the surface, that is just a nonsequitur.  Looking deeper down, this
argument starts from the assumption that there are only two
possibilities to consider: either give credit to *everyone*, or ignore
the issue of credit entirely.  The intermediate alternatives have been
arbitrarily excluded.  If you don't agree with that arbitrary
decision, the argument has no force.

If you think it is proper to use a name that gives credit to those who
developed a system, but you think (as I do) that it is impractical to
give credit in that way to all the contributors, I suggest making a
list of them in order of decreasing priority.  Then you can give
credit to the first N of them, for whatever N you think is feasible.

The GNU Project is the system's principal developer, so GNU ought to
be first in the list.  So if you use N=1, you would call the system
"GNU".  I think it is better to choose N=2 and write "GNU/Linux".
This has two advantages: it distinguishes this system version from
*the* GNU system proper (which doesn't use Linux), and it gives Linus
a share of the credit.  (I would not want to ask people to *stop*
giving him credit; he deserves it well enough.)  Meanwhile, it is not
impractically long.