[15:37] one problem with scapy might be that we'd have to create / parse lease files… oooooooor, we ignore that perform dhcp_request() every time we're called? alternatively, a lease could be a dict for the entire class, where we store that we've been called with interface x, from class y, and received IP z; if interface x still has ip z, we do [15:37] nothing… [16:44] rharper PRs 171 170 and 168 are fixed for starting SRU testing. [16:47] meena robjo +1 on the network configuration refactor suggestion discussion. I need to spend a little more time on robjo's existing PR trying to piece together if I have any concerns about this approach. I agree that the line between distro and network modules is unclean/muddy and needs a bit of cleanup. I'll try to frame what my perspective is on the best path forward. [17:03] blackboxsw: Keep in mind the current pending PR does make no attempts at refactoring anything, it works within the given framework and basically shoehorns more distro knowledge into the sysconfig renderer [18:30] robjo: which it could do without, if we did what you propose? [18:31] meena: Yes