In some email I received from Robert Elz, sie wrote:
[...]
> It is
> just that attempting to invent a new (object filename independant) way of
> getting offsets into /dev/kmem just has to be an unproductive way to
> spend time.   If there is to be a change at all, let it be a significant
> step forward, not just another wart.

Are you aware of where Linux has gone with respect to procfs and the above
and if so, what are your thoughts about their approach to solving this
problem ?

For those that aren't aware of where they've gone, they now have files such
as /proc/net/route with which you can do "cat /proc/net/route" (see sample
output below) and get an ASCII (hex in columns that have headings) routing
table.  Another example is /proc/net/tcp will give you a list of all TCP
connections.  In each case the output is all ASCII.

This is quite a step away from what is used for NetBSD now, but I'm sure
there are opinions about it being "forward" too.

Darren

% cat /proc/net/route
Iface   Destination     Gateway         Flags   RefCnt  Use     Metric  Mask            MTU     Window  IRTT                                                       
eth1    0002A8C0        00000000        0001    0       0       0       00FFFFFF        0       0       0
eth1    0002A8C0        00000000        0001    0       0       0       00FFFFFF        0       0       0
eth0    0001A8C0        00000000        0001    0       0       0       00FFFFFF        0       0       0
eth0    0001A8C0        00000000        0001    0       0       0       00FFFFFF        0       0       0
lo      0000007F        00000000        0001    0       0       0       000000FF        0       0       0
eth1    00000000        0102A8C0        0003    0       0       0       00000000        0       0       0