Talk:Linux naming controversy

Need input

I believe this article says all that it needs to, but I could use plenty of input. I started a thread in the forum here. --Joshua David Williams 11:44, 12 April 2007 (CDT)

I would suggest starting out with a paragraph that describes the issue, and then gets into the history. We go over the creation of GNU first, then Linux, then the controversy, and by that point, people are like "huh?". I mean, maybe if it started as "The GNU/Linux naming controversy is a disagreement between the FSF and various Linux groups about...", it'd appeal to people who are just dropping by to answer a question they had. I mean, in the case of a short article, the reader is likely here because they got linked here from another GNU and/or Linux page internally or from another source. They want a concise, two line answer with support and details. We're giving them details and telling them "mentally summarize it yourself". I wrote on the forums that it sounds like a lecture, and that's what it is, it's us giving details and letting the reader put the facts together. --ZachPruckowski (Speak to me) 12:13, 12 April 2007 (CDT)

We have conflicting edits. Here's what I wrote:

The GNU/Linux controversy is an argument held among many software enthusiasts of what to refer to a Linux operating system as. Technically, 'Linux' is just a kernel, that is, the core of the system. When Linux was first developed, the GNU toolkit was generally placed on top of the kernel to form a complete system. In order to attribute credit to the GNU development team, the Free Software Foundation (FSF) asks that Linux systems containing their software to be referred to as GNU/Linux systems. Not everyone is in favor of this convention, however. Perhaps the most common arguments against it include that it's unnecessarily long and harder to say, and that the Free Software Foundation is attempting to take credit for the work of the Linux kernel development team.

Which should we use? --Joshua David Williams 12:29, 12 April 2007 (CDT)

I think I like yours better. But one more thing; should we move this to GNU/Linux naming controversy? --Joshua David Williams 12:35, 12 April 2007 (CDT)

I think they're both good and we can use a blending of the two, or split and extend your paragraph into "Position of the GNU group" and "Position of Linux Advocates" sections. I mean, I think either work well. Let's see if others drop by in the next few hours and what they think. -- ZachPruckowski (Speak to me) 12:40, 12 April 2007 (CDT)

I think naming controversy might be superior, as there could be multiple controversies between GNU and Linux (GPLv3 stuff comes to mind, as do stances on binary-only drivers, etc). Let's see what others think, and let's look into how it's popularly used. -- ZachPruckowski (Speak to me) 12:40, 12 April 2007 (CDT)

Wikipedia uses GNU/Linux naming controversy. I think that's a far better name for it. --Joshua David Williams 12:43, 12 April 2007 (CDT)

Bold title

I'd suggest not putting the title in bold. See CZ:Article Mechanics#Bold titles ("the title of the article is an idiosyncratic phrase that does not name a single, particular item to be defined or briefly described") and my comment on CZ Talk:Article Mechanics. In fact, I would recommend starting the article with something more creative than "The GNU/Linux naming controversy is a ...". Fredrik Johansson 14:44, 12 April 2007 (CDT)

Linux advocates

I think the article is quite clear. I suspect the reason you are doing these pieces, Joshua, is as a natural extension of your Linux article. You should use this opportunity to discuss the controversy on the same Linux page, since you really, really can't describe Linux as an operating system without adressing this controversy. --Morten Juhl Johansen 06:11, 1 August 2007 (CDT)

Needs more balance...

This article is awfully light on information about the topic... It introduces the Linux and GNU project histories, but doesn't seem to say much of anything about the actual controversy/debate. It could really use some expansion.

In the intro, the justification for GNU-naming is mentioned, but no counter-argument to it is offered...

• I've often heard GNU-naming opponents point out that other open source apps like X11 can be considered more significant than all the combined GNU tools, by any metric (code size, frequency of use, availability of alternatives, etc.). So, X11/Linux?

• There is also the fact that the BSD utilities can be used as a complete replacement of the GNU tools, should anyone opt to do so. Similarly, the GNU tools can be used on a BSD system as easily as Linux. Additionally, there is much BSD code in Linux systems. So, BSD/Linux?

• Of course the counter-point to both is the importance of GCC; the one GNU project that is inescapably pervasive and currently irreplaceable, even in the BSD and proprietary-UNIX world. GNU/BSD? GNU/Solaris?

The justification of Linux-naming, for reasons of "simplicity," seem a pretty weak counter-point to stand on its own, when many much stronger reasons do exist. Ryan Cooley 11:42, 2 October 2008 (CDT)

Historical quibble

The article currently has "Before Linux, there was no choice but to use non-free kernels, most notably Minix, atop the free GNU utilities and tools." I don't think that is quite right. GNU tools were portable for 32-bit systems, but not 16-bit ones like many Minix boxes. Minix had its own tools including a compiler.

Here's an explanation of that history which I wrote elsewhere [1].

Then there's Linux, also out of academia but by a somewhat different route. In the early 80s most universities had Unix, including the source code, but the Unix license does not let you put Unix source on your slides or in your textbook. If you are teaching an operating systems course, you need code there. So a Dutch prof named Andrew Tanenbaum wrote Minix, basically a 7th Edition clone, that ran on PCs. His textbook, using Minix for all the examples, was very widely used. Minix was not free software, but you could buy the disks from his publisher.

The original PC and AT had 16-bit CPUs. Minix ran on those, or in 16-bit mode on Intel's 80386, and on the 32-bit Mac, but not in 32-bit mode on the 386. Tanenbaum showed little interest in fixing that; it would run just fine on a 386 in 16-bit mode, all you need in a teaching tool. Why make major changes to the code, especially if they mean you have to revise the book?

Nobody could distribute a 386 Minix; the publisher's copyright prevented that. What you could distribute were patch files that let anyone with a copy of the 16-bit version (presumably duly licensed) convert it to the 386 bit version. This was complicated and inconvenient in a bunch of ways.

Then a student doing a course based on Tanenbaum's book wrote a standalone 386 kernel that used the Minix file system and utilities. He announced it on the comp.os.minix newsgroup in 1991. Some of the reactions were along the lines of "Go away, kid. We're trying to do serious work here, getting these blasted patches to install", but other people reacted much better, in fact pitched in and started improving it. The kid was named Linus, so he called his kernel Linux.

Meanwhile, the Free Software Foundation had been developing a system called GNU (GNU's Not Unix), designed to be Unix-like and Posix-compliant but entirely free. Free in the sense of "free speech" not "free beer"; the issue is the user's freedom to modify, not cost. Their kernel project was bogged down in complexity, nowhere near ready, but they had a nice collection of other tools. These were designed for portability to more-or-less any 32-bit system, but most would not work on 16-bit systems.

Linux was a free 32-bit system. Add the GNU stuff (much of which the BSDs also use), some other things like the X Window system developed at MIT, and some borrowed BSD stuff. Presto; you've got a complete system.