I never quite got why GNU decided to consider the freedom to access the source code and the freedom to modify it as a single bullet point. To me, they're fundamentally distinct concepts, and furthermore the former is VASTLY more important than the latter.
Without the ability to check a program's source code, and personally compile it from that source, I am putting my hardware totally in the hands of the developer. I must trust that his program will do what it claims to do, and if it doesn't I can do absolutely nothing about it - other than threatening to take my business elsewhere or sue, which are merely the generic last resorts available for every kind of abuse or fraud. And even those will not shield me as long the program's misbehaviour goes undiscovered or unproven. It's pretty straightforward why this is a very dangerous situation when handling data of any importance.
However, if I can view and compile the source code, but not modify it, then I wouldn't say the loss of freedom is remotely comparable. If I see that the program contains objectionable code, then it is unfortunate and perhaps frustrating that my only option is to reject the program in toto, but I cannot see how I can legimately consider myself wronged. I was offered a product with all the information required to make an informed decision, the product simply wasn't good enough.
A food analogy: closed source is like having to buy food without the ability to examine its ingredients. Open-but-non-free source is like having access to the original recipe but not being allowed to take out the parts you are allergic to. It's a shame, but your health isn't in any danger.
I'm almost entirely in agreement with you, although I would say that the important distinction is between being able to access/modify the source code on one hand, and to redistribute it on the other. The right to access source and modify it, but not distribute, puts code in a similar category as books; certainly there is no restriction against making marginal notes or corrections in my own copy of a copywritten book, but putting the original or modified versions online for free download, or selling copies, would be illegal.
GNU seems to agree with you, since they categorised the right to redistribute on its own as freedoms 2 (original) and 3 (modified).
Of course, they also "consider [all non-free software] equally unethical", so from their point of view listing the four freedoms separately or as a whole should be pure cosmetics.
> I never quite got why GNU decided to consider the freedom to access the source code and the freedom to modify it as a single bullet point. To me, they're fundamentally distinct concepts, and furthermore the former is VASTLY more important than the latter.
They're fundamentally different for the purpose you're describing, but remember that your goal in this case is your goal. Stallman is protecting the freedom of software for any goal.
Besides, both freedoms are protected unequivocally without compromise by the definition. Saying the right to access the source code is more important than the right to modify it only matters if you plan to prioritize one over the other.
That's true, but I would argue it's counterbalanced by modern apps being (for the most part) always online. While we have more control over what data they can access, there are far worse things they can do with the data we do choose to share.
In the pre-internet days, I could write my sensitive documents in a closed-source word processor and pretty much the worst thing it could do was silently corrupt or alter my file. Now, a sufficiently malicious word processor could very easily upload a copy to its developer's servers, and make such activity nearly impossible to detect as long as it had a legitimate purpose to phone home.
Without the ability to check a program's source code, and personally compile it from that source, I am putting my hardware totally in the hands of the developer. I must trust that his program will do what it claims to do, and if it doesn't I can do absolutely nothing about it - other than threatening to take my business elsewhere or sue, which are merely the generic last resorts available for every kind of abuse or fraud. And even those will not shield me as long the program's misbehaviour goes undiscovered or unproven. It's pretty straightforward why this is a very dangerous situation when handling data of any importance.
However, if I can view and compile the source code, but not modify it, then I wouldn't say the loss of freedom is remotely comparable. If I see that the program contains objectionable code, then it is unfortunate and perhaps frustrating that my only option is to reject the program in toto, but I cannot see how I can legimately consider myself wronged. I was offered a product with all the information required to make an informed decision, the product simply wasn't good enough.
A food analogy: closed source is like having to buy food without the ability to examine its ingredients. Open-but-non-free source is like having access to the original recipe but not being allowed to take out the parts you are allergic to. It's a shame, but your health isn't in any danger.