> >I am also very aware of the profit-motive for >excluding open-source versions of drivers and apps. > >Even as a private user I have wasted hundreds of >hours trying to get hardware products to work only >to be told by the manufacturer that they *chose* >to refuse Linux access to minimal info. necessary >to write their own drivers. This anti-competitive >(on the software side) conduct is well-documented. >
It's not anti-competitive, and it's not sponsored or driven by Microsoft. It's just good business sense when viewed form the perspective of the hardware vendors. Hardware vendors are *notoriously* guarded about the details of how their hardware works. This includes the register-level details necessary to write a driver. And this is true for all types of devices that range from support chipsets (such as Southbus/ICH type devices) to commodity peripheral devices (like, say, SCSI adapters). They view their hardware interface as confidential and proprietary -- Sometimes because they think the interface design provides a competitive advantage, sometimes because as soon as an interface design is "public" they're stuck supporting every detail of it, and sometimes because they don't want their competitors to create register-compatible knock-offs of their devices. The VENDORS are the ones who write the drivers. They choose which O/Ses to write drivers for based on a cost/benefit analysis, taking into account one-time cost for writing the code and the on-going costs of supporting it. Vendors aren't any more forthcoming with Microsoft when it comes to details of how their hardware works. In fact, most typically they're down-right paranoid about it. So, if there aren't a plethora of drivers some flavor of Linux, it's just that the hardware vendor doesn't think writing and supporting a driver for that platform is worth the cost. There's really nothing more to it than that. Microsoft has been involved in a lot of conspiracies and anti-competitive practices, but this isn't one of them. de Peter K1PGV
