[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

(usagi-users 00022) Re: (usagi-core 00227) Re: [ANN] 1st release of USAGI IPv6 environment



At Mon, 6 Nov 2000 23:02:23 +0200 (EET),
Aki M Laukkanen <amlaukka@xxxxxxxxxxxxxx> wrote:

Hello,

> > > http://vesper.tky.hut.fi/mip/
> > That's great. If possoble, would you implement your MobileIPv6 on our
> > stack or cooperate with us ?
> 
> There's a small misunderstanding. University of Helsinki and Helsinki
> University of Technology are two separate entities although of course
> quite close. So it's not actually us implementing Mobile IPv6. You might
> want to contact them directly.

Oh, I got it. Thanks.
We will try to contact them.

> > > You seem to have evaluated the Linux stack with some kind of evaluation
> > > suite (pointer?). However the kernel used was 2.2.15. Have you run the
> > > same tests on 2.4.0-testxx? 
> > Not yet tested. But we have a plan to evaluate our 2.4 kernel and router
> > functions in the near future.
> 
> I think it should look a lot better. Alexey has done wonders with the
> code - at least with a quick look to 2.2 code.

Yes. We will evaluate 2.4-kernel ASAP.

> > Thanks. We have applied your patch into our repositry.
> 
> These small bug fixes are likely to get quickly in the main kernel
> tree too. What do you see as the role of you cvs tree vs. Dave
> Miller's cvs tree at http://vger.samba.org/ or the Linus Torvalds
> kernel? Personally, I'd prefer that cvs tree used for larger
> changes such as new features which can not be integrated until
> stabilised.

We consider that our CVS tree is the trunk of our new IPv6 stack.
It means that we will make larger changes such as replacing existence
codes or write from scratch. After evaluating our new code, we will
make effort to merge our patches into original Linux code.

> The mechanics of Linux kernel development very much prefer
> small, incremental patches. And I know Dave as the network area
> maintainer does so too. Many projects have not worked as good
> as they would otherwise if they had cooperated more closely
> with the main kernel; GGI, Donald Becker's network drivers etc.
> For example I am not sure that a relation such as which KAME 
> project has with *BSDs can work as good in Linux world.

Hmm. Of cource we will continue to make effort to merge our codes into
original codes because we don't want to make an another implementation
of Linux. 

Well, we sent our glibc patch to glibc maintainers yesterday and they
may include it on glibc-2.2.1.

Thanks for your comment.

-- Yuji Sekiya