[Dnsmasq-discuss] Removing a feature does anyone rely on this?

[Raymond Lillard]

On 07/01/2012 01:54 PM, Simon Kelley wrote:
> A quote from the man page:
>
> dhcp-range may have an interface name supplied as "interface:<interface-
> name>". The semantics if this are as follows: For  DHCP, if any other
> dhcp-range exists _without_ an interface name, then the interface name
> is  ignored  and  and  dnsmasq behaves  as  if  the interface parts did
> not exist, otherwise DHCP is only provided to interfaces mentioned in
> dhcp-range declarations. For DNS, if there are no --interface or
> --listen-address flags, behaviour is  unchanged  by  the  interface
> part.  If  either  of these flags are present, the interfaces mentioned
> in dhcp-ranges are added to the set which get DNS service.
>
>
> This feature was added to facilitate integration with uses such as
> libvirt, where libvirt could automatically add extra facilities to a
> single "system" dnsmasq instance. It has never been used as such, and
> things like libvirt and openstack and networkmanager have instead gone
> down the route of running their own "private" dnsmasq instances. This is
> not without its own problems, but we are well on the way to solving those.
>
> Since this feature is really rather odd, possibly misconcieved, and has
> not ever been used for its intended purpose, I'm considering removing it
> from future releases of dnsmasq. But if there is anyone using it for
> other than it's intended use, that's going to cause them problems. The
> best I can do to find out if this is used is to ask on the list. Not
> perfect, but if no-one here is using it, that's a reasonable indication
> that it's un-used.
>
>
> So, anyone want to step forward?

No complaints from me.

I am using dnsmasq in two installations for both
DNS & DHCP, including some of the more esoteric
DHCP options.  I can't imagine any cases where
the aforementioned feature is useful to me.

Ray