Thomas Bushnell BSG [u] wrote on 24/09/2004 18:31:
Adam McKenna <adam@flounder.net> writes:There are many ways for one to determine whether an IP address is dynamically or statically assigned. If you want to know more, do some research. Read NANOG-L or inet-access-L. It's all outside the scope of this list.Geez, you keep saying this. Surely if there are many ways you could give me just one? I really do know, I really have done the research; I've been using and programming IP for twenty years. So this is, from an expect, a put-up-or-shut-up demand. If you know a single reliable method, say it now.
In case anybody else would wonder:Adam did _not_ mean anything like actual remote detection of wether an IP address was assigned by DHCP, but rather those DUL-(black)lists and/or certain name parts in the address's PTR record.
Really, very smart (not) of him not to rectify this misunderstanding here on the list.
There doesn't seem to be any _reliable_ method of finding out wether an IP address is assigned dynamically (or even detect wether it is assigned by DHCP) short of asking the ISP (either through his reports to certain DUL IP-lists, analyzing the PTR record or asking directly).
Problem is: Except for asking the ISP directly about a certain IP, these methods aren't really reliable. For one, the DUL lists contain many address spaces besides those (hosted and) reported by ISPs. Secondly, quite a few ISPs assign IP addresses from a "dynamic" subnet statically for a variety of reasons (though a similar variety of reasons make this a bad idea, mostly because of those stupid DUL blacklists). And finally, there are many IPs assigned dynamically (or used for NAT, which is somewhat equivalent) but not listed in DUL blacklist. That last problem is no problem to me, but it still exists.
Anyhow, I think we all have made our points clear, and this discussion wandered pretty much OT, so this is my last post on the subject unless there is a very good reason to post more (I don't see that coming though), so if anyone would like to discuss all this in more detail, I am happy to set up a seperate mailinglist for it or discuss off-list.
cu, sven