Linux-libre project meets rocky reception

[rickh]

http://www.linux.com/feature/142772

if Alexandre Oliva, a Red Hat employee best-known as a board member of the Free Software Foundation Latin America (FSFLA), has his way, building a 100% free distribution will become easier thanks to his linux-libre project. Unfortunately, the path to freedom, he's finding, is often blocked by politics and a preference for convenience over ideals.

I hope Debian is more receptive to this project than Redhat/Fedora.

[BioTube]

I think a lot of people are afraid that a 100% free kernel would eventually knock the normal kernel out of the distro, making them either have to work harder for their hardware or compile a kernel on their own.

[rickh]

I think a lot of people are afraid that a 100% free kernel would eventually knock the normal kernel out of the distro, making them either have to work harder for their hardware or compile a kernel on their own.

Small price for freedom I would say. (Like bolting down the seat.)

[GNU.Wasabi]

Mmm... What benefits are there for end-users of a 1000% free kernel? As in what kind of license will it be?

[MeanDean]

uh...the same benefits that free software provides...

[BioTube]

I personally believe in the old axiom.

If it ain't broke, don't fix it.

[EMD]

I think a lot of people are afraid that a 100% free kernel would eventually knock the normal kernel out of the distro, making them either have to work harder for their hardware or compile a kernel on their own.

Mmm... What benefits are there for end-users of a 1000% free kernel? As in what kind of license will it be?

uh...the same benefits that free software provides...

I agree with both Dean and BioTube on this one.

Providing a 100% free kernel and forcing users to add non-free modules if they want sound, wireless, etc. sheds light on hardware that requires non-free drivers. (I wouldn't have known that some of my hardware requires proprietary drivers if I hadn't had to build the modules).

People who want to run 100% free software should be empowered to do so (and not be encumbered by other people's proprietary hardware).

That having been said, people who need proprietary drivers should have an easy way to install them (e.g. using Module Assistant with packages from Debian's non-free repositories).
.

[MeanDean]

If it ain't broke, don't fix it.

thats the problem....having closed code in the kernel is some broke ass crap...

[dmn_clown]

I hope Debian is more receptive to this project than Redhat/Fedora.

You're forgetting N.I.H. syndrome that prevails in FOSS.

[kinematic]

There are GNU/Linux users who -believe it or not- use their computer as a tool. Some may even depend on it to get work done and don't have the time to mess around with non-free drivers or getting to know all the intricate details of their hardware. While it's nice that some die hard FOSS advocates fill some of their spare time with campaining they also have to look at the useability side of things.

[EMD]

If it ain't broke, don't fix it.

thats the problem....having closed code in the kernel is some broke ass crap...

When BioTube used the term "broke," he meant "non-functional." Obviously, one cannot say that all of the proprietary kernel drivers are non-functional. To the contrary, they frequently provide the only way some users can obtain functional sound and wireless.

There are GNU/Linux users who -believe it or not- use their computer as a tool. Some may even depend on it to get work done and don't have the time to mess around with non-free drivers or getting to know all the intricate details of their hardware. While it's nice that some die hard FOSS advocates fill some of their spare time with campaining they also have to look at the useability side of things.

Producing and distributing proprietary software is not immoral per se, but proprietary software creates an environment in which a lot of immoral behavior can occur and that immorality often renders proprietary products useless.

For example, at my previous job, I noticed that the proprietary statistical software package that my boss had just purchased (SPSS) was incapable of computing a chi-square statistic properly. Under normal circumstances, I would have contacted the company for support. Perhaps they would have responded with a fix or perhaps they would have been immoral and declared that the bug is a feature.

Unfortunately, circumstances quickly became abnormal and I never had a chance to contact the company. My boss promptly informed me that SPSS is high-grade, professional software, she informed me her software doesn't make any mistakes and she further informed me that I would be fired if I ever checked for accuracy again.

(She must have spent a lot of money on that package!).

By contrast, if you ever notice an error in one of R's functions, simply type the name of the function without any arguments or parenthesis. That will output the text of the function, which you can modify and then reinsert.

@ kinematic -- Which statistical software package would you consider to be more usable? The proprietary package (SPSS) or the Free Software package (R)?



At the end of the day, it is more sensible, ethical and practical to run Free Software.

That's not to say that I haven't considered the arguments of the opponents to the Free Software movement. Most of the arguments against Free Software that I've found tend to be ill-informed. (A lot of people think that the Free Software is a Communist plot to destroy good American businesses).

I did however find one response to a pro-Free Software blogger which asserts that Free Software is promoted by some businesses to undercut their competitors. Even if that were true, it doesn't change the fact that the marginal cost of producing and distributing software is zero dollars and zero cents. Competition in the software industry should push the price of software to zero.

To survive businesses will have to sell complementary products and services, like installations, customizations training and support.

In the case of device manufacturers, they already sell a complementary product (the device itself), so what sense does it make to keep the software a secret?
.

[BioTube]

Your boss is evil. Did you report the bug anonymously? And I'm not against a free kernel, just being forced to use one. The installer could ask, though(with a "If unsure say 'non-free'" note).

[EMD]

Your boss is evil. Did you report the bug anonymously?

I spent the next month struggling to keep my job. Reporting the bug was the last thing on my mind. I ultimately left and found a better job where my hard work and attention to detail are appreciated.

And I'm not against a free kernel, just being forced to use one. The installer could ask, though(with a "If unsure say 'non-free'" note).

That's why Debian should provide modules for the non-free drivers that Debian has removed from the kernels.
.

[Jackiebrown]

There are GNU/Linux users who -believe it or not- use their computer as a tool. Some may even depend on it to get work done and don't have the time to mess around with non-free drivers or getting to know all the intricate details of their hardware. While it's nice that some die hard FOSS advocates fill some of their spare time with campaining they also have to look at the useability side of things.

What people don't seem to remember is that Linux would be next to OS2 if it weren't for it's dedication to free software.

If Nvidia decides to stop supporting my video card, I would not be able to use my computer as a tool...

Proprietary software and hardware puts us forever dependent on a manufactuar.

[FolkTheory]

i wish the best to this project. i wondered why FSF doesn't use Debian and i guess its something related to kernel blobs like this but oh well. anyways, best of luck

[katofiad]

There are GNU/Linux users who -believe it or not- use their computer as a tool. Some may even depend on it to get work done and don't have the time to mess around with non-free drivers or getting to know all the intricate details of their hardware. While it's nice that some die hard FOSS advocates fill some of their spare time with campaining they also have to look at the useability side of things.

The ability to study and modify is integral to usability.
Even if a proprietary driver works today with the current stable kernel, denying the kernel developers driver source prevents them from maintaining that driver in future kernels. Once the proprietary driver breaks a GNU/Linux user loses every advantage that free software would otherwise have provided. They are back to a situation where their only recourse is to beg the developer, because they have surrendered their freedom and given power over their computer to a third party. They cannot fix any problem themselves, and if they are not a programmer, they cannot even pay a programmer to fix the problem for them. Even if the driver works today, its flexibility and maintainability are hopelessly broken. The only way to avoid this situation is to specifically buy hardware whose drivers are free. This discourages companies who make hardware from keeping drivers secret, and gives them an incentive to release specifications. Then everyone at every level of development is empowered to fix problems or add improvements and upload a patch for everyone's benefit, from the kernel team right down to the end user. Those who continue buy hardware whose drivers are non-free aren't only foolishly throwing away their freedoms, they are paying others to continue to treat the freedom of other people with disrespect.
It is a vital necessity in free software development to require the freedom to run, study, modify, and redistribute every program, with source code, in order to ensure the most usability to every user right now and in the future, as well as deny any possibility of third parties gaining power and control over the users.

[ninjatux]

People who want to run 100% free software should be empowered to do so (and not be encumbered by other people's proprietary hardware).

You make it sound as it's really a service to develop a free kernel, when I imagine most people run some proprietary hardware.

I think this is really a pointless debate. Why are Linux users so caught up in being "free"? I love F/OSS, but the focus of F/OSS is to do what at least one proprietary vendor hasn't, which is to provide outstanding software that performs well, is secure and stable. And, even though Linux comes close, I haven't seen it come anywhere as close as other Unix/Unix-like operating systems. So, why sit here and argue about this? I imagine that a lot of developers waste their time with things like this too.

By the way, I like the idea of sharing, but my computer is specifically a tool that allows me to complete tasks the way I want to. My only requirement is Unix. I do not work well on Windows because of my computing style. I don't care if the Unix I'm using is proprietary or not. I care about getting my work done. I also don't care to spread a particular ideology either, even though I do appreciate F/OSS a lot and participate in mailing lists. I'm of the BSD mindset that a person should have the right to decide how to distribute their code and or modifications to existing code. I don't particularly appreciate the GPL because I think it's a bit limiting.

[katofiad]

I'm of the BSD mindset that a person should have the right to decide how to distribute their code and or modifications to existing code. I don't particularly appreciate the GPL because I think it's a bit limiting.

The BSD license is good enough to capture some level of community invention, but it does so sub-optimally. The problem with non-copyleft free software licenses is that they do not clearly define to those who are investing their time and/or money in a project what the split is going to be down the road. The code could end up in proprietary hands where it will disappear and make lots of money for their competitors. It could remain in sharing hands who return contributions. Or it could fork out of all possible control and create a chaos of multiple invention. Historically, BSD'ed projects have taken all of these paths. Copyleft licenses guarantee to the investor that the code will at the very least avoid usurpation by proprietary interests and at best gather community support that donate value and user innovation to the project. That is by no definition any limitation to anyone other then those who wish to be able to study and modify other author's work and use their modifications to gain power and control over the users of the software.

[Jackiebrown]

Why are Linux users so caught up in being "free"? I love F/OSS, but the focus of F/OSS is to do what at least one proprietary vendor hasn't, which is to provide outstanding software that performs well, is secure and stable.

I am not sure where to start with that one.

Don't mistake your motives with those of F/OSS

[MeanDean]

So, why sit here and argue about this?

No idea....