On Aug 25, 2009, at 4:49 PM, Nicolas Williams wrote:
On Tue, Aug 25, 2009 at 04:50:42PM -0400, Dale Ghent wrote:
On Aug 25, 2009, at 4:22 PM, Nicolas Williams wrote:
On Tue, Aug 25, 2009 at 01:34:11PM -0400, Dale Ghent wrote:
http://elektronkind.org/outbox/afs/openafs-s10u4-compat.patch
What's with the use of IN_CLASS[ABC]() and IN_CLASS[ABC]_NET?
IIRC, and it has been some years since I've had work work in this
code, is that it's used in part for determining if a AFS server is on
the local subnet and/or if a AFS server is local, on the same host.
Don't you want to use the _real_ netmask? CIDR deprecated the
notion of
classes A, B and C way back in 1992! :)
It seems lot of what AFS uses and does was deprecated way back in
1992 :( If you want to talk old and crusty, there's still a healthy
amount of K&R style in there, but that's getting cleaned up now the
last I heard.
My goal with that patch was to just get the client working again on
S10U4 and later, and a lot of the logic in that patch is duplicated
from what was there before - just the source of the information was
different. But that's when I worked at a .edu and making use of AFS
was part of the daily grind... now, not so much
/dale
_______________________________________________
networking-discuss mailing list
[email protected]
