Re: RTLinux patent
On Mon, 23 Oct 2000, Raul Miller wrote:
> On Mon, Oct 23, 2000 at 02:09:09PM -0600, John Galt wrote:
> > The modifications would probably be so great that you'd have a
> > copyrightable work in and of itself.
>
> I'm not convinced, yet.
We're talking about the basic scheduling of the kernel. This is so low a
level within the kernel that the divergences between minix, mach, and
386BSD are veritable chasms compared to the binary incompatibilities. I
was going to provide some examples, but the the plain and simple fact of
the matter is that the diff file would be OTO the size of the kernel
sources, and none of them leap out as the obvious place where a little
tweek here and a patch there would make them close enough to use the code
from RTLinux as a guide. I suppose that you could use the RTLinux as a
guide to making a RTFBSD (I happen to have a freeBSD box handy for
reference) if you really wanted, but you'd probably have been better off
in this case writing it from scratch.
> > However, as Marcus has pointed out, the algorithms are the sticky
> > issue and that would be enough to consign it to non-free if it was
> > distributable at all--Debian-bsd seems to be back from the dead,
> > and the *BSD kernel (which one to start with is still a matter of
> > contention) is most definitely NOT GPL :)
>
> Of course, this only is significant for third world countries (e.g. the
> u.s.a., maybe) which allow patents on such architectural issues.
>
> > AFACT, the rule of thumb is that if it's not distributable with any
> > part of Debian, it's not distributable by Debian PERIOD, which is the
> > way IMHO it should be.
>
> I presume here you're talking about copyright -- we have linux specific
> binaries which don't work under hurd, for example -- you just plain
There is a difference between "not working" and "not able to ever try to
get working". That difference is DFSG 6.
> can't distribute working hurd versions because they don't exist. Also,
> we have non-us software which for patent reasons isn't distributable
> for systems which exist in the u.s.
<sarc>Yeah, like mp3 encoders...</sarc> In the mp3
figh^H^H^H^Hdiscussion, it was mentioned that a mp3 encoder could fit
nicely in non-US. This was shot down rather quickly with the explanation
that non-US is not for avoiding US patents, but for avoiding US export
controls. Ther have been countless proposals to split non-US into
patent-controlled (like RSAREF used to be) and export-controlled
(like RSAREF is now) sections, but their resounding failure can only mean
that the split is moot because Debian doesn't deal with patent contolled
stuff (mp3s, GIFs, ad nauseam) unless there's an overriding need, which
RTLinux doesn't fall into.
> > So I'm going to weasel out of your question by saying that it's
> > irrelevant in this case :)
>
> Ok, but I'm going to weasel you right back -- guard your socks.
>
>
--
Customer: "I'm running Windows '98" Tech: "Yes." Customer:
"My computer isn't working now." Tech: "Yes, you said that."
Who is John Galt? galt@inconnu.isu.edu, that's who!
Reply to: