Voice of the Masses: Is GNU still relevant?
Although our magazine is called Linux Voice, we will give plenty of credit to GNU and RMS for getting the Free Software movement started and providing much of the software that makes up the GNU/Linux system. But a question needs to be asked: just how relevant is GNU (as a software project, not a movement like the FSF) today?
Some distributions are talking about moving from GCC to LLVM/Clang, thereby jettisoning one of GNU’s flagship programs. If this happens, and the GNU contributions to GNU/Linux are merely smaller components like Coreutils, should we really still keep talking about GNU? Or will GNU always be relevant because of its history? Let us know your thoughts for our next podcast!
Related Posts
About The Author
Mike Saunders
Mike stores his data by printing out hex dumps and laminating the sheets. He's the author of The Haynes Linux Manual and the creator of MikeOS, an operating system hand-built from assembly language.
Yes. GNU is much more then just technology, it also represents a very much progressive philosophy of sharing and protecting our digital freedoms and rights, which are being eroded so fast in todays times. So I would say that today it is even more important to talk about this philosophy that GNU and FSF represent and educate people about how important it is. SO this is why GNU is very much relevant if not even more relevant today.
+1 here.
As a philosopher I am genuinely puzzled why the issues GNU aren't debated more, and my current view is that the problem is twofold: political and PR.
The USA is a trendsetter in global culture and thought, with a historically founded political liberalism throughout. This is constitutive for what it is valid to talk about.
The GNU projects and FSF are academically founded projects, and may raise concerns that don't fall within the restrictions of political liberalism.
Which brings us to the 2nd item; the communicating of this distinction is often overlooked in GNUs and the FSFs communications.
We need GNU licenses and software, like the US needed the French Revolution for its founding idea to exist.
The technology is also subject to meritocratic measures, being a geek means optimization, but the GNU compiler and coreutils exist as Free alternatives, and justify philosophical scrutinization of projects that choose non-free or socially reckless licenses.
It is indeed confusing to a European why someone would choose a non-Free license for their projects in the US, when the the country is wrought with social differences, a smaller middle class than South American countries, poverty, and private interests infecting the political system.
It is not a question of whether the GNU project and FSF should exist, but why their message isn't communicated properly outside academic circles, when the need is greater than ever.
Wot that other dude named AntonT said…
Most people out there already don't know what the GNU in GNU/Linux stands for, I think we owe it to the likes of RMS and others to educate new comers to the free software community what we owe these pioneers of free software, and that without their
original work we would not have the Free OS and other software that we use on a daily basis.
GNU has these days more and more users because they are rediscovering what they lost some time ago : the four essential freedoms
(0) to run the program,
(1) to study and change the program in source code form,
(2) to redistribute exact copies,
(3) to distribute modified versions.
We need GNU philosophy like we need The Universal Declaration of Human Rights.
GNU is about freedom and how much we appreciate it.
When the user said is a GNU user you know he/she appreciate the freedom.
If a user said is a Linux user its just said it appreciate convenience with closed source code in kernel and programs all over the places with closed source code with Digital Restrictions Management , anti-features, spyware.
Convenience is what Windows and Apple does and they offers digital handcuff.
Before we dismiss GNU (and before I back your project) — I'm curious: what do you think GNU is?
If you see it as just GCC + coreutils, I think there's a lot you're missing…
Of course it's more than GCC and Coreutils. It's a project, a philosophy, and a movement. Still, GCC is arguably the biggest software contribution to GNU/Linux distros today, and if it gets replaced by non-GNU software, that affects the GNU-ishness of the OS we use. Hence why it's useful to debate these things…
What about stuff like GNOME? 🙂
One thing I like about LXF is that they're a corporate patron of the FSF. Any plans there?
Exactly, GNOME is a good point. A lot of people don't know that it's a GNU project… That's my idea with this discussion. Pretty much everyone agrees that the FSF does very important work, but I'd like to hear opinions about how relevant GNU (as a software project) is. So I've tweaked the wording a bit.
Meanwhile, I'd love for Linux Voice to be a corporate patron of the FSF. But ultimately it's our readers who decide where the profits will go:
http://www.linuxvoice.com/giving-profits-back-where-and-how/
Lots of support for the FSF there, so it looks very likely to happen. We'll be giving away so much more and getting more deeply involved than the old magazine.
It's a lot more software other than gcc and coreutils:
http://en.wikipedia.org/wiki/List_of_GNU_packages
Well, that's a silly question.
You mention coreutils and GCC, but you're forgetting one very important component: glibc. If you're thinking, "oh, but that's just the C library, who cares about that?" you must be ignorant. Otherwise, you will recognize that the fact that several otherwise completely different operating systems use the same implementation of the C library makes them binary-compatible. It's because they all use glibc that you can run the same binary on GNU (if it's ever released), Trisquel, Ubuntu, Fedora, Arch, Parabola, and Slackware.
Unfortunately, it's popular to call GNU binaries "Linux binaries", which leads to confusion; these binaries don't work on Android, which also uses Linux to talk to the hardware, but they will work on GNU or the Debian variant that uses the GNU Hurd, which don't use Linux.
In fact, to most users, GNU matters a lot more than Linux, because Linux only matters for hardware compatibility; once you have that hardware compatibility, as a user, you don't tend to notice any improvements to the kernel, so one kernel that works on your hardware is virtually indistinguishable from another kernel that works on your hardware. But the fact that several systems are GNU systems means you can grab a GNU binary and run it on your system with confidence, whether you have Fedora, Slackware, Trisquel, Dragora, or some other system based on GNU.
So of course GNU is still relevant. It's just not widely recognized, while the relevance of Linux is blown out of proportion in people's heads.
I'm not sure you understand what a C library does:
"It's because they all use glibc that you can run the same binary on GNU (if it's ever released), Trisquel, Ubuntu, Fedora, Arch, Parabola, and Slackware."
"GNU" will be using The HURD. It will not run Linux binaries natively, even if it uses Glibc:
http://www.gnu.org/software/hurd/faq/binary_compatibility.html
I would argue GNU coreutils is very relevant. It is what I use most to talk to a computer. Many Unix sysadmins will install GNU. Even if they don't much proprietary unix software requires GNU to be available to run. When you use unix you get used to putting a g in front of every command to get it to work as expected. Can't uncompress that archive? try gtar. Relink not working? Try gmake.
Gnu is the principled backbone of OSS.
The Spine.
Without that solid spine behind him, Torvalds wouldn't have the power to raise the finger at those who deserve it.
He would have been forced to live his life too deep in their pockets to dare.
We need the principles of Gnu and the pragmatism of Torvalds.
We need both.
Steps like this are also done with the kernel where a distribution is using a different kernel with a GNU based userland. Debian has had a freebsd based distribution for some years now. Does that mean Linux is becoming less relevant? If experimenting with a different compiler means GNU loses relevance, then actions like replacing the entire kerne would mean that the linux kernel is losing relevance even faster.
See
http://en.wikipedia.org/wiki/Debian_GNU/kFreeBSD
Agree with all the GNU support earlier. Now, what about the most fundamental element the GNU GPL! That's as significant a contribution or more than anything else!
+1 GNU… very important
I dare say all the movement towards BSD/Apache licenses can be attributed in a large part to Apple or developers who used to use Linux and now rely on a MacBook (oh it's ok I use linux on servers… and how does that help people not using servers?!). Apple seem to go into Anaphylactic Shock at the mere mention of the GPL, even hiring the CUPS dev to make sure it never went GPLv3 and not shipping Samba anymore above v2 (instead their own half baked solution). While they play along with FOSS when it suits they aren't Free Software's friend. It would be a real shame if the GNU for all it's faults was made less relivent, it's more than just a license it's all the previous posters mentioned issues too.
If KDE wasn't GPL and Apple hadn't been made (initially they were forced to make changes) to contribute back to KHTML which became Webkit would we have had the likes of Chrome etc. and a more open and complete web? Firefox was stagnating a bit at the time and if there was competition so early on with Google throwing in a 3rd browser would it have worked? I can only hope the LLVM competition spurs on the GCC folks and leads to a resurgance (can always hope!). BSD code in the longrun, I beleive, is "selfish". Sure you as a dev/user, right now, have a free piece of code, but it doesn't protect the user 2, 3 or more steps down the line when someone takes it in house makes it proprietary and more shiny and everyone uses that because they can't be bothered any more.
People seem to be only too happy to give up once something is shiny enough… (I include myself here as well… although I do try… mostly 😉
We need single minded, stubborn people to get the last few steps. That always seems to be GNU.
It's interesting that the GNU/FSF fanboys immediately start howling about philosophy and principles when the question was about relevancy as a software development project. I am not educated enough in the inner workings of Linux/GNU(sic) to even venture a guess about GNU's current technical contributions, but I do find this an interesting question. Some unbiased answers would be great.
If you use a GNU/Linux(sic) system, you almost certainly make
direct or indirect use of most of these GNU packages amongst
others:
GNU GRUB
GNU Bash
GNU Coreutils
GNU Sed
GNU Awk
GNU Grep
GNU Binutils
GCC
GNU Make
GNU Autotools
GNU Gzip
GNU Tar
GNU Nano
GNU Readline
GNU Wget
GNU Ghostscript
GnuPG
GDB
GNU Findutils
GNU Patch
GLib
GnuTLS
GTK+
GNU C Library
GNU Parted
GNU Diffutils
GNU Ed
GNU Groff
GNU M4
GNOME
GIMP
GNU Emacs
GNU Time
GNU Aspell
GNU Screen
If you subscribe to FLOSS mailing lists then you almost certainly
use GNU Mailman.
Finally, in this age of massive online surveillance, there are
several very important relevant projects under the GNU umbrella. These
include the already mentioned GnuPG and GnuTLS but also GNU
MediaGoblin, GNU Social, GNU FM and GNUnet.
So yes, it seems to me that GNU is very relevant as a software
project.
I think the problem is you can't base a project purely on technical merit anymore. There will always be a potential license issue, patent troll or take down just around the corner. Or someone trying to make a proprietary project based on open code but attempting to lock it down. I think the philosophy has become part of the technical contribution at this point. Although the FSF has had some ham-fisted implementations of it in the past.
The Samba team being prime example of making sure CIFS was accessible to all.
I think "yes" is the answer to your question 🙂
Avoiding the well-put philosophical points, It's just another component of Linux – so you just can't avoid it (esp, if it's prefixed as GNU/Linux 🙂 )
All distros use different components as others, but I expect they'll all have a GNU component in there somewhere, so you'll *have* to discuss GNU anyway…
So, the answer's "yes" 🙂
Hmm…is GNU still relevant…I dunno, has anyone tried installing a distro completely free of Gnu stuff and using it. That would probably answer this question.
No opinion on the GNU project as a whole but the growth of LLVM and Clang is a very worrying trend. LLVM has a really innovative architecture from a technical standpoint and if it does eclipse GCC as the mainstream C/C++ compiler platform it will be a sad day, as commercial interests, a self-selecting group of which who dislike the GPL, are controlling its development. Accepting LLVM/Clang is a first step towards nullifying the GPL as a concept, which is what these companies want while paying lip service to 'open source'.
I still can't find a good a reason to replace GCC. I have several HPC tests and LLVM/Clang just don't seem worth it. I agree with David; the GPL is the most important piece of literature in our world. It IS the software I love and it is the philosophy behind every decision I make in my life. GNU/Linux isn't about being the best technically it is about making the world a better place. The GPL is the Constitution for a Free Digital Life. Had GNU Software not been there, I doubt Linux could even have been compiled in the first place.
The voice of Linux is GNU, Linux may be Bull, but the GNU Herd are the people behind it!