Guy Fraser wrote: > Rather than promote fear, why not lobby to get the dbkm or dkbm > what ever it is that will make dynamic binary kernel module The reasoning is, why accept to be locked out of sources for the modules when the module exists because of the donation of work that generated the kernel itself openly? People who sank a lot of work in the kernel become grumpy when they see that work applied to devices that cannot really be owned and controlled by the user. > system available. If developers have a static API to design > there modules, there could be more drivers available. Some > devices manufacturers have secrets they are legal bound to > protect, that prohibit them from releasing source, and it is > a fact of life. If you keep kicking the manufactures in the The manufacturers accept such restrictions only because they know a binary-only distribution is acceptable, for example, in the Windows world. At the edges already Linux challenges this assumption (because the benefits using it brings outweigh the risk from loss of obfuscation of the code) and can bring about change. So it isn't enough to consider the situation as it is, one also needs to consider what Linux can make be by its policies towards closed source modules. I don't buy the story that certain classes of device can never have open source drivers regardless of all considerations. If Linux is desirable enough to become a no questions design-in, then it will impact decisions all around it and that can be in a positive way. -Andy
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature
