On Wed, 3 Jan 2001, C. Stephen Gunn wrote:
> I agree that you could educate ifconfig in the ways of netgraph
> and hide it all behind the command interface you propose. It's a
> migration to a broader view of interfaces for ifconfig(8). Right
> now, ifconfig(8) is basically a front-end for ioctl()'s on a single
> network existing interface.
Well, with netgraph support it'd be ioctls plus netgraph's messages. I
don't think it'd be THAT incompatible or bloated compared to original
ifconfig.
> The UNIX paradigm is powerful because of many well-made, single-task
> tools. In most regards, ifconfig is complete. Adding significant
> functionalty causes ripples. For starters, libnetgraph moves into
> libstand, and picobsd. Or we could fork ifconfig(8) to have two
> variants.
Sometimes in the course of human events it is necessary to break with
compatibility ;) I believe netgraph is sufficiently advanced and well-made
system that it should be used as much as possible.
> I'm not saying it's wrong. I'm just saying that it makes me feel
> dirty and violated.
>
> </SOAPBOX>
>
> > Given their lack of netgraph, and apparent reluctance to implement it,
> > that doesn't seem much of a problem at this time.
>
> Yup. I'd prefer to see FreeBSD take the higher-road and strive to
> be compatible and cooperative whenver possible. Instead of continued
> isolation, divergance, and proprietization. Hey, I don't even
> currently run Net/OpenBSD. ;-)
There's nothing that would prevent OpenBSD people from taking netgraph and
implementing it. I run OpenBSD, and threw the idea a few times on their
mailing lists, and response was less than enthusiastic...Which lead me to
choose FreeBSD for my next networking project. :)
-alex
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-net" in the body of the message
