🚀 go-pugleaf

ruhl@4dv.net (Robert Uhl <ruhl@4dv.net>) writes:

> Kevin Everets <kevin.everets@alcatel.com> writes:
> > 
> > > When the FSF _demands_ this kind of "quasi-worship," it just gets
> > > peoples' backs up, and discourages them from expressing gratefulness
> > > where gratefulness is due.
> > 
> > This statement is false, I believe, and appears to be making the whole
> > thing more of a religious issue than a pragmatic one.
> 
> The logical statement `if an organisation demands quasi-worship, it
> gets back disrespect' is false?  Curious, that.

Sorry, I should have made that clear: the FSF does not demand any kind
of "quasi-worship" (that I've ever seen, at least).

> > From what I've read, Stallman's out to get credit for the GNU
> > system, and nothing else.
> 
> He wants the GNU label place over everything on a typical Linux
> distro, above TeX, above netbase &c.  He wants to elevate it to a
> level of importance equal to the kernel.

He seems to want the GNU label to be the label people use for a
typical GNU distro (which happens to use Linux) as the point of view
being presented is that GNU is the system.  It's not "elevated" above
the kernel.  Why should the kernel be elevated above the windowing
system, or the Desktop Environment?  The problem may stem from the GNU
Project naming the kernel something different from the system ("the
Hurd", as opposed to just "GNU's kernel").  Solaris, FreeBSD, and a
myriad of other examples didn't do that, so perhaps that's why they're
spared from this overly long thread?

Out of curiosity: if you took a Solaris system, and put the Linux
kernel in it, would you call that the Linux OS?  My guess is that you
would.

> > The GNU page on Linux seems to express fairly well the GNU project's
> > position on this: they were the first to begin the massive attempt
> > to develop a complete free Unix-like system.
> 
> And the first to fail in a very big and embarassing way...

I'm pretty sure that it's far from a failure... it's still under
development (and is in use).  I guess you have a different definition
of failure than I do.

> > None of those individual components are an attempt to create a full
> > operating system, however.
> 
> Non of those attempts are full OS, _and neither are the GNU tools_.

You're right: no single GNU tool is an attempt at a full OS.  A full
GNU OS includes (at least some of) them, though.  Why name a
collection by a single component?

> > Basically, if you don't believe that a system is any more than the
> > kernel, then by all means continue to call your system "Linux".  If
> > instead, you feel that a system is more than a single component then
> > what you're talking about is a version of GNU that happens to have
> > Linux at the helm.  Thus, it would be called "GNU/Linux".
> 
> In the latter case, what I'm talking about is the Linux OS with all
> the myriad bits of free software, GPLed, LGPLed, BSDLed &c. which make
> up my system.  Thus, I call it Linux.  Although I could, I suppose,
> call it BSD/MIT/GNU/Knuth/Linux.  Which would be silly, and dumb.

Calling it Linux/X/TeX/LibC would be dumb too... and that's why no one
is suggesting it.  You appear to be of the opinion that the kernel is
the most important piece of an operating system, and thus should be
the name of the operating system, which is well within your right to
do.

I would worry, however, about things like Lindows which by your
definition is the Linux OS.  If viruses get written for Lindows (being
as by default that Operating System runs as root, and so it would be
easy to do), then many lament that Linux will get a bad name.  If you
instead hold the position that Lindows is an Operating System in its'
own right (that happens to use Linux, many GNU programs, and some BSD
ones in addition to their own extensions), then there's no issue:
Lindows is insecure.  So what?  Use GNU and be happy.

> > Unless, of course, you're talking about a proper GNU system (simply
> > called "GNU") that uses GNU's official kernel: the Hurd (which is
> > actually quite usable, despite dubious reports to the contrary).
> 
> I'd call that a Hurd system.  I don't call my Solaris boxes with the
> GNU tools installed GNU/Solaris; I call 'em Solaris.  I wouldn't call
> a Hurd box with the GNU tools installed GNU/Hurd (I might refer to GNU
> Hurd, but that indicates something else).  So why should I call a
> Linux box with the GNU tools installed GNU/Linux?

This paragraph is redundant; you've stated your position that the
kernel is the most important piece of software and so should be the
only thing that names the system.  Why bother writing this out again?

> Answer: because someone's bitter that someone else could come up with
> a GPLed OS in half the time and get twice the performance.

You seem to be very bitter yourself, though I don't understand your
stake in it.  I don't perceive bitterness from the GNU project,
though.  Nor do I think that the time spent on the project or its'
performance is an issue to those who use GNU without Linux.  The
capabilities and hackability of a full GNU system seem more
important.

Another thought: I can at least see consistency in the arguments of
the "Linux OS" crowd as they seem to think that GNU Project == RMS
(the project as only one member of it, though that member is probably
the most important) and thus GNU OS == Linux (the project as only one
component of it).

Cheers,

Kevin.

-- 
Kevin Everets <kevin.everets@alcatel.com>
GPG: 0x86D6C5DD 48FA 053B FB7D 9BF9 FE4C  02D3 6346 055B 86D6 C5DD