Leon Brooks wrote: > On Fri, 9 Aug 2002 06:45, Igor Izyumin wrote: > >>Sometimes, they don't have a choice. > > Yeah, like about once in every two blue moons. > > >>They may have trade secrets or >>something within the driver that would prevent it from being open-sourced. >>For example, if nVidia open-sourced their driver, their competitors could >>use their work for their own chips. Do you think they want that? > > No, I don't. But it _is_ a totally bogus excuse. Their competitors _will_ be > clean-rooming their chips and disassembling their drivers anyway. It's one of > those if-you-outlaw-guns-then-only-outlaws-will-have-guns problems. As things > stand, *only* their competitors have access to their `secrets' and not you or > I, not their more-or-less friends! What could be a worse situation than that? > Better to also give their allies access, no? > > Some point-haired-boss moron lawyer makes that decision, not someone with > their brains actually operating. > > Cheers; Leon > >
IMHO, the fact is that sometimes the same company don't probably totally owns the code they include in the closed (or partially closed source) drivers. they ship. E.g. for NVidia I've read it's due to code written from SGI (GLX). Also other companies have claimed the same policy (e.g. the Matrox has the HAL library in closed source object binary only [and thus not included in XFree]). Others instead tried a different approach and let their drivers to be included open source in the main linux/XFree tree (e.g. Adaptec, Compaq for RAID). In other cases, instead companies released binary only drivers but abandoned it (e.g. the ESS module for ESS based winmodems, or the Lexmark printer drivers) so they would work only on old kernels/distributions. Bye. Giuseppe.
