that old GNU/Linux argument

[Björn Persson]

Ric Moore wrote:
> So, he took Minux, an operating system, (presumably with it's GNU
> packages), inserted his kernel (which took only 10 months to have it
> working enough to distribute), named and released a package named
> "Linux", under the GPL. He referred to it as "an operating system". Now,
> either Linus understood what an operating system meant, or he was just
> ignorant and misguided as hell.

There is no agreement on what "operating system" means. Some, like Andrew 
Tanenbaum, say it's pretty much equivalent to "kernel". By that definition 
Linux is an operating system. Others, like Richard Stallman, say an operating 
system contains all the programs you need to use the computer to perform 
various common tasks. By that definition Linux and GNU together form an 
operating system.

> Is Minix now GNU/Minix as well?

Andrew Tanenbaum is entitled to choose the name of what he produces and 
distributes. If he has packaged GNU tools together with his own kernel and 
distributes that package, then he can call the whole package "Minix", just 
like the Fedora team call their distribution "Fedora". The GNU tools are then 
part of Minix just like they are part of Fedora.

If the end user who receives Minix from Andrew Tanenbaum has to get the GNU 
tools separately from the FSF and combine them himself to form a usable 
system, then it would be wrong to say that the GNU tools are part of Minix.

Now, which is the case with Linux? Do Linus Torvalds and his team distribute a 
complete usable system, or do they distribute a kernel that needs to be 
combined with the GNU tools (or equivalents) to form a usable system?

Björn Persson