IPv6 posting backoff patches for CURRENT snapshot

[Jeffrey M. Vinocur]

On Sat, 6 Apr 2002, Antonio Querubin wrote:

> Although it was known that the posting backoff routines never worked for
> IPv6 clients in the CURRENT snapshot, after looking through the source I
> have to wonder how it ever worked for IPv4 clients as well.  Attached are
> patches that hopefully make it work for both IPv4 and IPv6.  

I don't use the posting backoff, but after some small changes (it's silly
to use sscanf() to parse the dotted quad when inet_aton() has already done
it) it compiles cleanly.  If anybody actually uses this, let me know if I 
broke anything.


> The attached patches also incorporate some other patches for nnrpd I
> submitted earlier in the week but which haven't made their way into the
> CURRENT snapshot yet.

By the way, it's best if you can avoid doing this -- we prefer to commit 
separate patches separately, so I just end up separating them out again.


Tony, thanks a lot for this; almost all of the outstanding IPv6 issues are
now fixed.  So Russ, the only thing worth holding up the next release for
is the need to make some file descriptors into arrays in innd (Katsuhiro
was going to get to this eventually).

-- 
Jeffrey M. Vinocur
jeff at litech.org