What if we made software with the expectation that end users would make at least a few changes? This would greatly simplify things.
Though they don't really say so in their philosophy statement, I think the suckless.org software tends to be like this. Configuration/customization of their software tends to be: change the code and/or header files, and recompile.
It makes the code simpler and reduces the attack surface area by reducing the handling of untrusted input.
To me there is a slight difference between open source software and editable software.
Open source is editable, certainly, but most users will never do this. They will install a package from their package manager and use it, exactly as they would install an app from the App Store. The software is written with this expectation.
Editable software would be written with the explicit expectation that a user will tweak it to his or her liking. It would be like buying unfinished furniture or a roughed-in house -- the framework is all there, but the buyer is expected to do some work to get it to its final desired state.
Though they don't really say so in their philosophy statement, I think the suckless.org software tends to be like this. Configuration/customization of their software tends to be: change the code and/or header files, and recompile.
It makes the code simpler and reduces the attack surface area by reducing the handling of untrusted input.
To me there is a slight difference between open source software and editable software.
Open source is editable, certainly, but most users will never do this. They will install a package from their package manager and use it, exactly as they would install an app from the App Store. The software is written with this expectation.
Editable software would be written with the explicit expectation that a user will tweak it to his or her liking. It would be like buying unfinished furniture or a roughed-in house -- the framework is all there, but the buyer is expected to do some work to get it to its final desired state.