On Tue, 2006-02-14 at 17:29 -0600, Mike McCarty wrote: > Craig White wrote: > > On Tue, 2006-02-14 at 13:32 -0600, Mike McCarty wrote: > > > >>Les Mikesell wrote: > >> > >>>No, that should be a different topic: How the GPL sabotages open > >>>source adoption. The GPL restrictions have done more to > >>>maintain the Microsoft monopoly than any Microsoft employee. > >>>Even when a vendor tries their best to supply drivers it > >>>doesn't work out very well and they aren't accepted in the > >>>distributions. > >> > >>Well said, Les. > > > > ---- > > there are many forms of open source licenses that aren't GPL and a > > vendor can supply drivers using a form more suitable to their liking if > > they wish...I guess I missed Les' point. > > Well, you need to read the GPL and LGPL more closely. > In order to produce a runnable program (or a driver) > one needs to link with libraries. These libraries > are either GPL or LGPL, makes not much difference. > Both GPL and LGPL have virus-like properties of infecting > all the source for the whole program, causing the producers > to reveal their special interface. So, they can't legally > hide the code which knows trade secrets about the hardware. > > To keep the hardware proprietary, they can't reveal the > program source, and they can't link the object without revealing > the source, so they don't produce a driver for Linux. > > I hope that wasn't too turgid. ---- not at all and IANAP so you clearly know more about this than I but I thought 'link' only applied if it were statically linked and thus, a driver file itself didn't represent an infringement of GPL/LGPL licenses, only the manner which they were used might be. Craig
