How FSF (and free software zealots) miss the point of Free Software

My physic teacher used to say:
why is a religious question. Actually the real important question is how; since once you understand the mechanism of something you can improve it. While wondering about the finality don't bring any useful answers that can be neither checked nor used to influence the system we observe.

This is my first attack on FSF: while I love the definition of free software given by FSF, I am very reluctant to adopt the ideological views of FSF.

Just to refresh memories, the definition of free software by the FSF is made of 4 freedoms that should be granted by the licence: non-discriminative usage, sharing, studying, modifying the code that should be provided when a software is released. It results in a single important property of these software: they can be forked.

This definition is accepted by everyone and is the reference (even for OSI), and some weired spirits says that the FSF licence is a tinge limiting the freedom of usage compared to others licences such as BSD.

The main distinction between Open Source and FSF (gnu project) is the finality.

Open source is considered a pragmatic or materialistic approach where softwares are viewed as economical externalities for which the cost of developing is so high given the available resources that sharing is just a mean to make the cost decrease while increasing the «frontier» of new problems to solve.
FSF thinks software is about freedom. The bits of code we share according to them have much more power: there are the foundations of a new free society to empower citizens. They are the partisans of an «intelligent design» of software.

For them, we face a competition of the bad proprietary software trying to enclose us in a technological lock-in due to the very nature of economy based on externalities; according to Chicago school, software/OS industry should tend towards natural monopolies: the more you use my goods the more my costs diminishes thus the more I will win money even if my product is crap as long as it is adopted.
FSF also thinks citizens can be free if they have tools to express themselves and see computer networks as the climax of the modern Gutenberg press.

To sum up: Free software is for the FSF a synonym for free speeches, and free society.

So in order to avoid the lock-in religious zealots focus on trying to provide alternative to «potential lock-in technologies» needed to build an independant functioning OS:
- kernel (Gnu HURD ;) );
- system (GNU bash, openSIP, GNU...);
- development (GCC, Glibc, Gnu ADA, mono, Guava, Gnu ...); 
- security (GNUTls, GPG...);
- desktop (Gnome..);
- office suite (Gnumeric :) );
It is exact to say actual linux distributions are using in order to be functional a lot of GNU technologies brought up by the FSF. Thus the claim of FSF one should not say Linux but GNU/Linux (the pronunciation is available in .au somewhere on the fsf sites and worths a good laugh).

While we live under the empire of necessity (externalities), it is wrong to praise the necessity; there is no necessity to leave under any empire may it be from the forces of «right». We should totally consider getting rid of most of the FSF sponsored software when they are harmful.

And that is my point of opposition with FSF. I kind of agree with their view, I strongly disagree with their way of trying to achieve their goals that for me is counter productive in terms of engineering and of education.

Hence the how vs the why approach.
FSF code alternatives to lock-in technologies without wondering if:
  1. they are competent in their field;
  2. these technologies are beneficial from the beginning.
GNUTls considered harmful, rants about how clueless the TLS team is about C coding and maintaining code. Since security is kind of like walking in a very dense minefield when the code looks like behaving like a drunken man from the most simple point of view of coding, then you don't trust the code.

FSF zealots will say: not a problem, by the shear property of openness and amelioration the code will tends towards better code.

Well, this is wishful thinking, crappy engineering even with good QA very rarely tends to be good engineering at the end. (bf110, fulmar ...)

And kaboom you walk on a mine: a bug in GNUtls made it possible for 5 years to bypass certificate checking. The NSA really must fear FSF claims that being convinced of freeing society makes software that rocks.

And still FSF is making a FUD (fear uncertainty doubt argument) on proprietary software dangers promoting «safer» free software... Please!  

Is GNU-tls an isolated case? Well, Gnome (for using
c# notably everywhere), gnumeric, gcc, glibc, mono... received a lot of criticizes for their engineering, and there are more. As Linus Torvalds says: all bugs can become security bugs, so the first rule of security is to adopt correct software.

Plus, some alternative are even worse than no atlernatives at all.

Office suite bringing to the computers all the confusion of mixing what you mean and what you see, and the stupid "paper" analogy of documents is harmful. People should focus on the content. Software are gifted for applying a lot of stupid rules: they are gifted for versioning, applying templates, typographic rules, having hyperlink, access control in a distributed environment. And we still use this bloatwares called office suites making you write on a virtual piece of dumb paper.

Our document papers designed by computers are when you know the rules of typography below what we can do with manual typesetting. Typographic rules are not for grumpy old men, they are a way to higher the speed of reading and comprehension of written documents. Yet with these awesome computers, we have documents that are pathetically less readable compared to what we could do.

FSF is by definition reactionary and conservative by «proposing alternatives» to adopted lock-in technologies.

Their ideological blindness make them back up wrong so called «new technologies» like sheeps. Maybe TLS is wrong. Maybe C is wrong, maybe traditions are wrong.  Maybe they are right. But I am sure blindly reacting to «proprietary lock-in» by quasi systematically proposing free software alternative is dumb. It sometimes helps the adoption of incorrect technologies.

Thus and here is my conclusion: FSF is harmful.

Against all evidences Free software zealots and fanatics are using the post Snowden era as a way to advocate GNU/free software. Saying that by property because people that do code are «good people» validated by a Political Kommissar inquiring their views they do «good software».

Well, fuck no.

openssl (which is not GNU), gnutls are below average crypto suites, with very harsh engineering critics since the snowden revelations.

I don't mean the proprietary alternatives inspire me much confidence (like RSA's stuff).

I say, we don't want «good or free» softwares. We want softwares that are well built and thus we can trust.

FSF says open source by enabling the «proprietarization» of software is evil.

Well, before windows adopted the BSD stack, their TCP/IP stack was much more vulnerable to sequence prediction attacks. They may have changed their stack since windows NT.
But what I can say is it is better the TCP/IP communication between to computer be safe. TCP/IP don't care about linux or windows or BSD, and one compromised computer make a lot of people unsecure.

So once more, I prefer windows to use open source software that is well engineered because it also selfishly helps me being safer.

And last and least: what matters in a software is not what it is said to do or any phantasmagorical values, but its correctness. FSF is just doing marketing for its own chapel chapel to which I don't belong in my name, and I strongly oppose it.

I am a dysfunctional small part of free software yes. I could even be rated a failure: I totally can live with it. But, still, I am part of it.

FSF should bear in mind it doesn't own free software or its values. Even a very insignificant developer as I have divergent opinions than theirs, like a lot of devs that code instead of writing stupid blog posts as much as I do.

Defining correctly the four freedom of software magnificently does not give FSF the right to pretend speaking or expressing the view of free software communities.

Collaborating on something does not imply we share a common view. And that is the freedom zero of free software they forgot: the radically non discriminative freedom to use free software whatever your opinions are.

And for all of us that are just «using» softwares for pragmatic reasons of having «correct» tools, when they push towards unsafe «uncorrect» softwares using a FUD on security, there is no way they cannot piss us off a little bit.

We don't need a unified free software. We don't need political strength. We don't need a wider adoptions of «free software» in security for instance, we need a wider adoption of «correct» approaches to security that will not be possible if FSF enforces the adoption of poor technologies by providing even more broken alternatives to proprietary approaches (based sometimes on the open standard they cherish like oauth2.0).

We don't need alternative to the desktop «à la windows or apple» we need correct desktop.

We finally don't need more «adoptants» that want free software everywhere, we just need more educated people that can make enlightened choices not based on fear but on understanding. And it would be great if they helped us find new disruptive approaches based on really new technologies to solve the old legacy problems left by crappy softwares and design some of them coming from GNU softwares.

And I am bored about their lack of culture in computer history: the first community aiming for users to be able to grow their own «vendor independent solution» is not born in 1984 with Stallman but the SHARE user group in 1953.

Please, you don't get credit for modifying history. You just looks like an Orwelian dystopic movement, or acculturated religious trapped in their closed mindset.