pf/nat & PPTP _client_ question

[Pete Leonard <pete_(_at_)_hero_(_dot_)_com>]

Hey folks,

Setup:

OpenBSD as my home firewall/server.  All clients within are DHCP'd.

I have a work laptop that I use regularly at home.  To connect to the
office environment, it's PPTP (don't ask - trust me, I've asked for better
w/o luck..).

So I've got everything up & running under OpenBSD 3.1.  Removed GRE
support, added the following rules to the firewall:

pf.conf:
pass in  quick on $ExtIF proto gre all keep state
pass out quick on $ExtIF proto gre all keep state
pass in  quick on $ExtIF proto 53 all keep state
pass out quick on $ExtIF proto 53 all keep state

nat.conf:
rdr on $ExtIF proto gre from any to $ExtIF -> 192.168.7.33 
rdr on $ExtIF proto tcp from any to $ExtIF port 1723 -> 192.168.7.33

And everything works fine.  

Here's the question - Is it possible to re-configure this so that I don't
need to hardcode the 192.168.7.xxx IP address, but rather, have it adapt
so that I don't need to re-configure NAT every single time I connect a
client with a different IP address?

I've done plenty of searching on this - all examples seem to point to
active FTP & using ftp-proxy, or deal with the example of a VPN *server*
behind the firewall, running with a static IP.  Is the issue that some
kind of PPTP proxy would be required in this case?

What do people do when the FAQs aren't deep enough, the mail archives
aren't addressing your issue, and your expertise isn't extensive enough to
continue?

	thanks,

	--peter