On Thu, 2010-06-17 at 09:10 +0000, Antonio Gallo wrote: > For me the real question is: what is the brain? And then: is the brain > completely free? In our case the brain is the OS, that can ask the external > devices to do what he (the OS commanded by the user) desire. > > For me, it’s not a unacceptable to have a device that acts like a black and > stupid box, with its firmware installed directly inside him, that can do > simple things when someone ask him to do so (for example, send out an > electric signal when someone push a key). If the device can do only a limited > and well defined job, and if I have the freedom to tell the device to do > something when I desire, then the brain (the OS) is free. I can ask the > device to start/stop its job when I desire. I can’t modify the firmware of > the device to add more functionalities, but nobody can (apart from the > manufacturer if it could physically act upon the device). In this case the > firmware is not free software, but this doesn’t compromise my control upon > that device, and the OS is free software (even if the firmware of the device > is not free software). This means that if the functionalities named A, B and > C have been implemented in the device by means of the firmware > (directly installed into the device), functionalities can be controlled > through some different electric signals coming from the OS, then nothing can > prevent the OS to send, when requested by the user, those signals to the > firmware to order him to start the functionality A or B or C. This is because > the firmware+device is a black box programmed to response in a well defined > way when well defined inputs are passed to the firmware itself. You can’t > improve the firmware to add new functionalities (the firmware is not free > software), but you can totally control, when you desire, those > functionalities already implemented by sending electric signals to the > firmware (in this case the OS is free software even if the firmware is not). > In this situation (the firmware directly installed inside the device) the > manufacturer can’t act upon the device through the OS (if the OS has not any > blob installed). The firmware can't take any decision because it doesn't have > any element to distinguish. It could only take random decisions, such as > stop to work every 10 minutes, but independently from the OS that are > managing it. > > On the other side, if the firmware is loaded inside the kernel and if it is > proprietary and closed, then that unknown piece of code (program) can take > decisions from the inside of your OS and could choose, for example, to > manage/control the device depending on how you have configured your OS or on > how you use it. In that case the manufacturer can, in same way, control your > OS, the brain. It can choose to act differently on the basis of what it can > learn from your OS. > > For me the difference is subtle but very important. > > Bye > > Antonio >
Calling the brain an OS is inaccurate, as that would seem to imply a user that is external, a kind of ghost in the machine. The computational model of the brain is a brilliant one though, and almost certainly correct. It's not really directly related to this topic though, but thought I'd just point that out.
