On Fri, 2013-07-26 at 22:22 +0200, Andrew Shadura wrote:
> CIDR netmask is already supported. And yes, I thought about it, but
> what currently holds the implementation is that many hook scripts
> expect IF_ADDRESS variable to have just one address. Should I pass
> other addresses somehow else? Call scripts multiple times? Ask everyone
> to adjust them? I don't know.

Uhm depends,... does the kernel (and not just linux) somehow specially
tread the "first" address... ip show addr implies this somehow, by
calling the other "secondary"... if so... on could think of a special
stanza for the primary... but even then I sounds a bit ugly.
So probably having an "addresses" where all are specified and the first
one is the "primary" sounds best.

But how should one specify different broadcast addresses in such a
line? 

And should one integrate ipv6 into it?


There seem to be many further questions with the current syntax model.

Like where to ideally put stanzas like dns-search. Unlike the
nameservers this is often the same regardless of which interface is
used... so I add them to the loopback interface.



Cheers,
Chris.

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to