Re: writing a parallel port driver to support external hardware

Alan Cox (alan@cymru.net)
Tue, 30 Apr 1996 09:39:05 +0100 (BST)


> They're afraid of reverse enginneering, i.e. in a company somewhere in the
> intractable Far East (China, say) somebody looks at the driver's source
> code, has an immediate "ah, so that's how they do it" insight, proceeds to
> burn that into an ASIC or whatever, slightly modify your driver so as to
> spew out bogus company information, and hey, they're in business. (Are you
> seriously thinking about proecuting a (mainland) Chinese company for
> copyright infringement?)

Well all these binary only drivers are just causing more people to
reverse the technology and in doing so gain more insight than would have
occured if an uncommented free driver exists. Novell's attitude has put
almost the entire Novell protocol into the public domain and caused GPL'd
Novell server clones for example.

> In most cases, this argument is bullshit. However, there are a few
> legitimate cases. The insanity of interfacing a tape drive to a parallel
> port via a floppy controller may be such a case -- after all, all component

The floppy controller tape interfaces are documented. Hence ftape.

Alan