Forty years of GNU

Linus's contribution was a new, collaborative method of development. GNU may have been about freedom but the development process was behind closed doors. By the mid 1990s development of one of their two flagship products, gcc (the other being emacs), had stalled at version 2.7.x and numerous forks were out there trying to fill the gap. Eventually one of those forks, egcs, became so successful that GNU abandoned their own gcc (then at a very buggy version 2.8) and "blessed" egcs as the new GCC (now uppercase) version 2.95.

RMS and GNU may have "codified" free software but Linus, and ESR via his "The cathedral and the bazar" article, showed the world what the benefits really are.

That brings me to the second point. RMS tried to promote free software by haranguing, calling proprietary software evil and unethical, and so on. Predictably, that didn't work well. The "open source" folks promoted its benefits for development, collaboration, stability, importantly pointing out the benefits to the developers and corporations as well as the users. Their definition of "open source" was essentially identical to the FSF's definition of "free software", but it was persuasive where the FSF was dictatorial. And it worked. Yet, to this day, FSF disses the term "open source" on their website.

Linus changed the world, not just by writing a good kernel: he changed how people saw the process of software creation, even how GNU's own software is developed. (There were collaborative open-source projects before, but Linus's hierarchical stewarding was unique, I think. And git, which he wrote for his own needs, has become the standard VCS now, basically synonymous with version control.)

And the existence of Linux is what accelerated the development of almost all GNU software. This has been true for over a quarter of a century now. It is not a GNU operating system that happens to run a tiny kernel called Linux that could just as well have been Hurd. It is an entire ecosystem, including GNU and much else, that was developed mostly after the Linux boom, driven by Linux developers for the needs of Linux—I doubt very many of the 1980s-era GNU projects would still be alive if not for Linux, and most of the later projects would not exist at all.

And that is why the FSF's language in this release is so offensive.

Linus referenced GNU "just a hobby, won't be big and professional like gnu"

True, but it can reasonably be argued that that reference was already tongue-in-cheek. “Big and professional”? The GNU operating system mostly didn't even exist at the time – you certainly couldn't obtain a “GNU OS” tape and use that to install GNU on an otherwise empty computer –, but the FSF folks had already taken on board various other large pre-existing software packages such as X11 or TeX and summarily declared that they, too, were part of the GNU operating system. (Which was of course OK – these packages were very liberally licensed already – but it still amounted, in a way, to taking credit for stuff that they basically found lying around in the street.)

In the early 1990s, the “GNU operating system” was “big and professional” only as far as people's claims for what it might eventually at some unspecified point in the future be or do were concerned. For everyone else, GNU was mostly fantasy, and it took the efforts of Linus and his collaborators to make it reality to a point where the FSF could claim that what people could actually download or buy, not from the FSF itself of course but from entirely unrelated outfits like Red Hat or SUSE or Slackware, and install on an otherwise empty computer, was in fact the celebrated GNU operating system, plus a small and insignificant addendum called “Linux” – a fiction that they've kept going for the last three decades or so. (Oh, and let's not forget that according to the FSF at the time, Linux as a kernel for GNU was originally only a compromise, meant to tide us over until that most famous piece of vapourware, HURD, became ready, which would happen Real Soon Now, and GNU would finally come into its own. 30 years later we're not holding our breath.)

The Linux kernel became a part of “the GNU operating system” in the same way that X11 and TeX, among countless other components provided by non-FSF projects and developers, had become a part of it. So somehow today we're supposed to celebrate the FSF for “creating” GNU, even though the amount of code in GNU that was actually written under the auspices of the FSF forms only a minuscule part of the whole, and over its 40-year history the FSF has done precious little towards integrating the various components of “GNU/Linux”, making it installable, and generally turning it into a viable product – in other words, the drudge work. Writing lofty philosophical essays is one thing, and of course the influence of the GPL is not to be disregarded. In the end, however, the drudge work, overwhelmingly contributed by people and companies all over the world with no official connection to the FSF, and certainly very little actual overall leadership or technical direction from the FSF, surely deserves most of the credit for the ongoing success of “GNU/Linux”.