Re: "wireguard" implementation improperly merged and needs revert

["Jason A. Donenfeld" <Jason%zx2c4.com@localhost>]

Hi Nia,

On Mon, Aug 24, 2020 at 5:57 PM nia <nia%netbsd.org@localhost> wrote:
>
> Hi Jason,
>
> > We still need to protect the unique identity and reputation of
> > WireGuard (our "brand"). This ensures that when people see the
> > WireGuard name or logo, they know it is something we, the
> > WireGuard developers, have worked on."
>
> Personally, I would be in favour of entirely rebranding the NetBSD
> implementation to avoid this, because it's only introducing ourselves
> to potential legal problems.
>
> It's important that the NetBSD tree remains as free as possible, and
> that nobody introduce themselves to potential legal pain by modifying
> any part of it.
>
> We have also had a similar discussion with Mozilla's lawyers and
> simply opted to unbrand all of their software which we distribute.
>
> If only certain people can develop implementations of this protocol,
> this is not an open protocol.

Please read the reply I just wrote to Maya on tech-net:
https://mail-index.netbsd.org/tech-net/2020/08/24/msg007855.html where
I basically covered this already. I think you've been misled by
others' comments into somehow thinking this is related to trademark
stuff, when it isn't at all. And this isn't a situation with
"Mozilla's lawyers" or something either; there's no comparison, simply
because a "trademark" is simply not part of this discussion here --
i.e. were the question to come up when we're all ready to go here
about NetBSD using the name WireGuard to describe its implementation,
the answer would be an "of course" and if the question were then, "can
you put that in writing?" the answer would be, "yea, sure, why not."
This also doesn't have anything to do with "who implements the
protocol". Rather, the issue here is that NetBSD doesn't actually
implement WireGuard and its protocol. There's a lot more to WireGuard
than just crafting some packets that sometimes have the right crypto.
I'm afraid that Ozaki-san's code has been picked up with too much
haste and not enough study, and we're going to get into an ugly
situation if we don't put the breaks on now, before expectations run
too high, and reevaluate/restudy.

And just to put this discussion back into perspective, I *like* the
NetBSD project and I *want* to have everything work as smoothly as
possible, and I'm volunteering my *own development time* into helping
to make that happen. All I'm asking is that this trajectory here is
slowed so that we can do it right. Because I care about getting it
right.

Jason