Re: Turning GNU into a bottom-up organization

[Alexandre François Garreau]

Le jeudi 24 octobre 2019, 03:55:02 CEST Colby Russell a écrit :

> - Seven Laws of Sane Personal Computing <http://www.loper-os.org/?p=284>

>

> (Note that the latter is actually even more stringent, making GNU/FSF

> seem lax by comparison, and all on purely technical grounds; not even a

> machine that runs libreboot at power on to boot directly into GNU Emacs

> on a linux-libre kernel would satisfy the principles laid out in the

> Loper OS charter.)

 

It is an interesting reading, and some of these laws might look seducing at first, but all these “shall” (though local and in context) leads to confusion, for me. As well as your statement about universal absence of software complying with it (why talking about GNU not doing it then?). It is way different from software freedom [0].

 

Most of these issues are technical, and not to follow them is an imperfection, a technical stance. Totally orthogonal from software freedom[1]: proprietary software is made to attack your freedom. “Unsane” personal computing isn’t.

 

This is well expressed by the problem of someone accidentally obfuscating “preferred form for edition of a software”: that might be a social problem, but it’s not oppression as we can’t stop people from “refusing” to be bad at programming, at explaining themselves, etc. It is also possible for someone to very easily write uncommented software in assembly, that will be impossible to understand, that will be the case most of the time (and it would even to the author then: that’s the biggest difference from proprietary software, since author doesn’t have as much power anymore[2])… basing on such software or making it important would be —as well socially— a bad idea, but it’s not wrong, we don’t ought to eliminate this, to forbid it (by the very rule “forbid to forbid”), because it is not “forbidding”: it is not a supplementary act you do that weaken others’ freedom, that you could instead not do, it is not here to do that.

 

Also this then only apply to generalistic computers, not to embedded computers (while most of those of course have end users (thus software freedom problems still apply)).

 

Moreover, law I has been said to be difficult to implement (freedom —at least socially— is not difficult to implement, it is power which is wrong to implement); law II has been said to be, in its extreme form, inpractical (especially on low storage) and then disruptive (it would complexify usage and interrupt thought flow for sometimes negligible good: it though stays a good non-extreme idea to keep in mind); law IV goes against low-level debugging, necessary to develop compilers, hardware, etc. which may be operator’s purpose, and is necessary for a metacircular fully-bootstrapable system; law V (as well as law III and other laws actally) prohibit user arbitrary trading between features (interactivity and resources usage); law VI goes against law IV (actually in roman numeration it’s coincidentally the reverse…) and doesn’t work if resources are too low compared to initial state; law VII is against user interest to know when they break social rules, so they do it better if they want.

 

All that, that shall be doable with free software (because freedom #0 (well it says you shouldn’t be socially prevented from doing it, not from breaking some theorical contract on the internet)), show that this is merely technical discussion, not social nor political.

 

[0]: I’d never thought I’d so vehemently cite that article I knew by disagreeing: https://www.gnu.org/philosophy/imperfection-isnt-oppression.html

 

[1]: note with a simple implied “all right reserved”, proprietary software could still do “sane personal computing”. Because this only speak about technical possibilities, not social one. It doesn’t speak about law enforcement.

 

[2]: this is enlightning. Developer of proprietary software have power because proprietariness gives them power, arbitrary one. They can keep it (without ending loosing it), they can delegate it, transfer it, etc. Because it is a social power (because within GNU philosophy, freedom and power are defined as social). It is not something you loose by accident. It is something provided by the society, a society of enslavement[3], not a free society[4]. When some developer is just bad at writing, explaining, collaborating, etc. he can loose it, he cannot delegate it, transfer it, etc. thus his influence, his ability to change the environment (his “inner power”) is way more limited than with social power given by proprietary software, because it is individual. He cannot conspire with others to limit freedom. He doesn’t steal freedom by having power, doing harm. He merely do lesser good to the society.

 

[3]: how would better “non-freedom” be said in english without unwanted connotations, like “asservissement” in french?

 

[4]: as in the say “free software, free society”