> > Really? Here's an example where this is not true (anymore?).
>
> I was referring to this [1] implementation which makes a minimal use
> of BSD features -- at least when I ported it.
>
> [1] http://www.awfulhak.org/ppp.html
Okay, that clarifies a lot of things ;-)
Regards,
-Farid.
--
Fari
> Really? Here's an example where this is not true (anymore?).
I was referring to this [1] implementation which makes a minimal use
of BSD features -- at least when I ported it.
[1] http://www.awfulhak.org/ppp.html
___
Bug-hurd mailing list
[EMAIL PRO
> And the tunnel device has all the semantics from the BSD tunnel device,
> because that was my reference implementation at that time.
The only semantic that um-pppd relies on with respect to the tunnel
device is that it acts like a bidirectional fifo.
___
> Really? Here's an example where this is not true (anymore?). In
> FreeBSD 4.5RC (-STABLE), last cvsupped 01/20/2001:
^
Of course, I meant 01/20/2002 ;-)
Sorry about the confusion.
-Farid.
--
Farid Hajji -- Unix Systems and Network Admin | P
> > BTW, a BSD-based pfinet is highly desired, mainly because of
> > um-pppd. Recent changes in the FreeBSD Net/3 (especially the
> > netgraph(4) infrastructure [e.g. ng_pppoe], KAME, ...) influenced
> > 'ppp' so much, that it would be very hard to synchronize FreeBSD ppp
> > with the Hurd versio
> On Wed, Jan 23, 2002 at 05:12:30PM -0500, Neal H Walfield wrote:
> > Wrong. The only reliance that um-pppd has on pfinet is a tunnel
> > device; everything else is pretty much standard.
>
> And the tunnel device has all the semantics from the BSD tunnel device,
> because that was my reference
> "Farid" == Farid Hajji <[EMAIL PROTECTED]> writes:
Farid> Look at netgraph(4) in FreeBSD:
Farid>
Farid>
http://www.FreeBSD.org/cgi/man.cgi?query=netgraph&apropos=0&sektion=4&manpath=FreeBSD+4.4-stable&format=html
Farid> It looks like an interesting and probably portable w
> > > 1). I propose to move all linux code out of pfinet and make separated
> > > library with linux ip stack.
> >
> > This is not a horrible idea, and I once thought of doing the same
> > thing myself.
>
> What would be *really* nice is a nice library with nice well-defined
> and reasonably s
[EMAIL PROTECTED] (Thomas Bushnell, BSG) writes:
> Peter Novodvorsky <[EMAIL PROTECTED]> writes:
>
> > 1). I propose to move all linux code out of pfinet and make separated
> > library with linux ip stack.
>
> This is not a horrible idea, and I once thought of doing the same
> thing myself.
Peter Novodvorsky <[EMAIL PROTECTED]> writes:
> So what can be done -- is to make some good interface in this
> library. It should generic enough for being applied to BSD ip stack
> too.
I don't think you understood what I wrote.
The details of what the Linux network stack expects from the rest
Hello, Jeroen!
> "Jeroen" == Jeroen Dekkers <[EMAIL PROTECTED]> writes:
Jeroen> On Sun, Jan 13, 2002 at 01:42:46AM +0300, Peter
Jeroen> Novodvorsky wrote:
>> I've read Protocol Service Decomposition for High-Performance
>> Networking and it influenced on me much. I think it r
Hi, Thomas!
> "Thomas" == Thomas Bushnell, BSG <[EMAIL PROTECTED]> writes:
Thomas> Peter Novodvorsky <[EMAIL PROTECTED]> writes:
>> 1). I propose to move all linux code out of pfinet and make
>> separated library with linux ip stack.
Thomas> This is not a horrible idea, and
On Sun, Jan 13, 2002 at 01:42:46AM +0300, Peter Novodvorsky wrote:
> I've read Protocol Service Decomposition for High-Performance
> Networking and it influenced on me much. I think it reasonable idea to
> insert a thin network interfaces code into kernel (actually what
> linux-src/net/core is). B
Peter Novodvorsky <[EMAIL PROTECTED]> writes:
> 1). I propose to move all linux code out of pfinet and make separated
> library with linux ip stack.
This is not a horrible idea, and I once thought of doing the same
thing myself.
However, two things block the idea:
1) The amount of Hurd speci
Hello!
I've read Protocol Service Decomposition for High-Performance
Networking and it influenced on me much. I think it reasonable idea to
insert a thin network interfaces code into kernel (actually what
linux-src/net/core is). But it is far goal and somebody might not
accept it.
One of ideas i
15 matches
Mail list logo