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

(usagi-users 00282) Fw: USAGI IPv6 patches

Discussuions between netdev ip maintainers and debian-ipv6 people.

-- 
Hideaki YOSHIFUJI @ USAGI Project  <yoshfuji@xxxxxxxxxxxxxx>
PGP5i FP: F731 6599 5EB2 BBA7 1515  1323 1806 A96F 5700 6B25
--- Begin Message --- 
Hello!

> I wondered if you could please tell me if there are any plans to merge
> the USAGI IPv6 patches[1] into the Linux source, 

Partially.


>					in order to make
> Linux comply with the latest IPv6 RFCs[2]?

It does.

USAGI patch contains some important things, but they are mixed
with of huge load of poorly cosmetic things and things, which
are difficult to accept.

Bug fixes are to be accepted. Unfortunately, it is not so easy
to separate really important pieces (I know only one to be honest:
reassembly of link local frames) of useless things sort of NI ICMP
and unadvertised change of macro names.

If USAGE people were more communicative at least about
core changes (bsdish binding rules) and real problems,
this would be great. Now it looks like a heap, where
it is not so easy to find real pearls.

Alexey


--  
To UNSUBSCRIBE, email to debian-ipv6-request@xxxxxxxxxxxxxxxx
with a subject of "unsubscribe". Trouble? Contact listmaster@xxxxxxxxxxxxxxxx

--- End Message ---
--- Begin Message --- 
(I!(BHola Alexey!

> >					in order to make
> > Linux comply with the latest IPv6 RFCs[2]?
> It does.

Partially. 


For the debian IPv6 project it's very important to have a good IPv6 stack
in mainstream linux kernel (because it'd be very hard to convince the
kernel package maintainer to use something else, and that's a good thing
IMHO)

> USAGI patch contains some important things, but they are mixed
> with of huge load of poorly cosmetic things and things, which
> are difficult to accept.

There are some things on USAGI that are very important for us (most glaring
probably is what USAGI names DOUBLE_BIND) and would like to see them in the
mainstream kernel.

> Bug fixes are to be accepted. Unfortunately, it is not so easy
> to separate really important pieces (I know only one to be honest:
> reassembly of link local frames) of useless things sort of NI ICMP
> and unadvertised change of macro names.

> If USAGE people were more communicative at least about
> core changes (bsdish binding rules) and real problems,
> this would be great. Now it looks like a heap, where
> it is not so easy to find real pearls.

Could we solve it like the Apache/SGI problem? Breaking the USAGI
differences in reasonably-sized patches and offering them to you?

Would you accept -for instance- a patch for the DOUBLE_BIND feature?
It's a real nuisance programming AF independant servers with the
mainstream linux behaviour. Yes, it probably breaks some IPv6 capable,
but not AF independant programs, but they'll be broken when the next
protocol arrives anyway...

> Alexey

					HoraPe
---
Horacio J. Peña
horape@xxxxxxxxxxxxxxxxx
horape@xxxxxxxxxx
bofh@xxxxxxxxxxxxxx
horape@xxxxxxxxxxx


--  
To UNSUBSCRIBE, email to debian-ipv6-request@xxxxxxxxxxxxxxxx
with a subject of "unsubscribe". Trouble? Contact listmaster@xxxxxxxxxxxxxxxx

--- End Message ---
--- Begin Message --- 
horape@xxxxxxxxxxxxxxxxxxxxxxxxxx writes:
 > Could we solve it like the Apache/SGI problem? Breaking the USAGI
 > differences in reasonably-sized patches and offering them to you?
 > 
 > Would you accept -for instance- a patch for the DOUBLE_BIND feature?

This is exactly what we are asking for.

Later,
David S. Miller
davem@xxxxxxxxxx


--  
To UNSUBSCRIBE, email to debian-ipv6-request@xxxxxxxxxxxxxxxx
with a subject of "unsubscribe". Trouble? Contact listmaster@xxxxxxxxxxxxxxxx

--- End Message ---
--- Begin Message --- 
(I!(BHola!

>  > Could we solve it like the Apache/SGI problem? Breaking the USAGI
>  > differences in reasonably-sized patches and offering them to you?

>  > Would you accept -for instance- a patch for the DOUBLE_BIND feature?
> This is exactly what we are asking for.

Great, i'll try to work on it tomorrow.

> Later,
> David S. Miller
> davem@xxxxxxxxxx

					HoraPe
---
Horacio J. Peña
horape@xxxxxxxxxxxxxxxxx
horape@xxxxxxxxxx
bofh@xxxxxxxxxxxxxx
horape@xxxxxxxxxxx


--  
To UNSUBSCRIBE, email to debian-ipv6-request@xxxxxxxxxxxxxxxx
with a subject of "unsubscribe". Trouble? Contact listmaster@xxxxxxxxxxxxxxxx

--- End Message ---