> Use Ekiga or an ordinary phone call.

Let's go over your list for an ordinary phone call. Here the 'client app' is your phone terminal, and the 'server' is the phone operator.

> - Is any communication not properly encrypted

Yes, it's not encrypted at all.

> - Is your communication being spied on / being recorded

I would say yes, at a similar level of probability as a skype call.

> - Is there a way to add feature "x" to the application

> - Is there a way to remove feature "x" from the application

The specs for the phone terminal are not open-source, so you can only do that with some hacking. The same is true for skype, even though there it might be more complicated technically.

> - Is there a way to use the application on your specific hard / software

Irrelevant for a physical phone.

The main difference between the skype client and a phone, is that a phone follows an open protocol, and therefore you can build your own. This is the only reason I can think of why using a phone is 'free' while using skype is not. But I think that if the skype protocol was made open, Stallman would still argue that it is not free and therefore should not be used, and therefore I do not understand his argument.

Right, but at least you can tell. No-one knows whether the encryption Skype claims to be applying actually works. And you can integrate an encryption system into a phone if you want to, whereas if you try and do that with skype they'll change the client so that your thing doesn't work any more.

> The specs for the phone terminal are not open-source, so you can only do that with some hacking.

Speak for yourself. There are open source phones available if you want them. There are no open-source skype clients.

> Irrelevant for a physical phone.

Sure. But I don't have to use a physical phone. I can run a virtual phone on my freebsd box, whereas I can't run skype.

> The main difference between the skype client and a phone, is that a phone follows an open protocol, and therefore you can build your own. This is the only reason I can think of why using a phone is 'free' while using skype is not.

That's not a small difference; I'd say it's the most important difference.

Stallman has always refused to use unfree software in devices he own or are directly using. Skype is within that criteria, while an Asterisk powered free software device is not.

So to understand the word freedom in the context of free software, I tend to lean back towards liberty as a philosophy question. What freedom means, is a question famous people have pondered over for ages. So to ask a single simple question if Skype gives freedom, here is a quote from John Locke:

"Persons have a right or liberty to not be subject to the inconstant, uncertain, unknown, and arbitrary wills of others."

Since the developers of skype can subject their users to any number of inconstant, uncertain, unknown, and arbitrary decisions, like including the NSA as secret listeners to private communications, Skype can not be defined as providing liberty.

For more details and guide to more reading: https://en.wikipedia.org/wiki/Liberty

It's not a software problem. Or at least, it's not purely a software problem. For the vast majority that freedom would make no difference. They don't know how to change things anyway. I remember a few months back on here when we were going 'Why don't people use PGP more? Oh yeah, the interface sucks.' And did anyone change the interface? Nope, not to my knowledge at least.

Did any of us make better encryption tools, despite there being available encryption libraries? Not to my knowledge.

That sort of freedom is tied to your capabilities as much as it is your tools. If your understanding of a computer goes something along the line of 'that magical glowing box under the desk.' it makes no difference either way.

To get to real freedom, in the sense that Stallman seems to mean, requires more than a change to open source software, it requires a system be built from the ground up to not just enable but also encourage co-evolutionary mastery. And it's not clear how that would work. Having the source code of the system available is probably a part of it, but you also need people to be able to start off with simple changes using the same language that the system is described in - preferably they'd use that language at least to some degree to instruct the system to do basic operations.

A lot of that probably has to do with how modular you can make a program - when you make changes how much of the program do you have to hold in your head to understand their effects? And I think that's probably, ultimately, a processor architecture level decision if you want it to be how most things on the system work.

Looking at it from the software side: at the moment, even with open source software, sure you can theoretically go in and learn about the program and make alterations and ... but that is not a trivial thing to do. Even for people with deep knowledge of computing it's not a trivial thing to go and learn about how the program works, and if they don't already know the language it's even worse. There are loads of open source projects out there at the moment where what needs to be done is essentially obvious but there's such a lack of people with the abilities and interest to do it.

Some people do use PGP, but they pay for it or it's an internal component in a larger system. GPG, the OSS version does have a sucky interface, but that's not the real issue people don't use it, there's three others:

1) Most people don't give a shit about encryption 2) Most people are too fucking lazy to understand/use encryption (Not being cynical, this is reality.) 3) Writing good encryption software is thankless and discouraging. There's a ton of commercial companies that use SSH and GPG that haven't donated squat to the developers.

"Did any of us make better encryption tools, despite there being available encryption libraries? Not to my knowledge."

You don't work where I do. I wrap GPG in another layer of software to make it more usable to the above average folks at my job. It's mainly limited to automated processes.

By their philosophy, it does. If that feels hand-wavey to you I would agree; I find the whole "define 'freedom' as 'things I want'" thing more than a little sneaky--but it is consistent with the rest of their philosophy.

My annoyance is simpler, though: I dislike propaganda. That the FSF seeks to re-define the word to the exclusion of all else is bothersome, much as, for another example, the American Republican Party seeks to redefine "socialist" to mean "center-right". It's the poor debate tactic of the extremist who can't accept the validity of opposing positions.