>>>> "sb" == SL Baur <steve(a)xemacs.org>
writes:
sb> Glynn Clements <glynn(a)sensei.co.uk> writes in
sb> xemacs-beta(a)xemacs.org:
> OTOH, does anyone know how standard the SIOCGIFCONF and
> SIOCGIFADDR ioctl()s are?
sb> That might be the best place to start. At least you can
sb> detect their presence or absence with configure.
> Also, what about IPv6 addresses (or even IPX, DECnet,
> AppleTalk, ...)?
sb> IPv6 numbers are a different size. DECnet uses small integers
sb> for node numbers. I don't know about IPX or AppleTalk.
On Linux, why not look at /proc/net/dev? That would give you realtime
information, it's standard on Linux no matter what user software you
might be using, it should get everything, and it's easy to access
(although find-file doesn't work on it---should it? if not, why not?
---`M-! cat' does).
One reason I suggest this is that we should probably be poking around
in /proc to see just how many system utilities could be simplified or
standardized on Linux by using it. Note that a lot of things can be
read from a running Linux kernel that way, some of which don't have
standard tools for access (eg the new kmod and IrDA subsystems).
That might almost justify the lack of portability to systems without
proc-fs.
--
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091
__________________________________________________________________________
__________________________________________________________________________
What are those two straight lines for? "Free software rules."