greenber@acf4.UUCP (ross m. greenberg) (01/19/85)
<>
Maybe this should go in net.wanted.sources, but I felt this was the
place. Flame me if I'm wrong :-)
I'm going to be doing some traveling shortly, and I want to be able to
foward my mail to wherever I happen to be. The cute trick is that
I don't know where this will be! I don't even know the sites yet.
I'll probably grub some sort of guest account along my journey!
(Anyone want to donate accounts to a worthy fellow in a line that
goes from London to Israel over six-eight weeks starting mid-Feb??)
I figured what I wanted was a program(s), or shell script(s) to
do the following:
1) Upon receipt of each mailgram (or periodically as per cron)
foward the mail to some site kept in a log file.
2) Upon receipt of a special message from me (remote) to me
(local), change the site in the logfile to reflect my new
site.
3) Send a message similiar to 2) to all sites that I've ever
send messages to as per 1).
4) Send the program(s) to all the sites in 3) in case
any mail for me is at that site. (perhaps just have it
send any mail to me at that site to me at "HOME" site?)
I guess that covers my wish list as of now. Of course the hard
part of this is the "special" message. Certainly I don't want
some nefarious guy sending a fake message to me and redirecting
all of my mail to him!! Now 1) seems pretty easy via cron and
not too easy for each mailgram (once or twice a day would probably
suffice, though).
Point 3) could be me just running the same program as in
1) entered manually as I leave a site, I suppose. But there is no guarentee
that I'll know where my next site will be!! So I'll hard code in
a return to HOME site as the initial log file, and wait for the message
from 2) to tell me where to "really" send future mailgrams.
It does feel sloppy, though. Just think about the turn-around time!!
I mean computers are supposed to help automate this type of stuff, right? :-)
One thing to consider:I am not root on this machine and don't
expect to be on the guest machines. So all this has to run non-privl'ed.
Does anybody have any thoughts on this? Anyone have code that already
comes close??
Send me mail, I'll summarize, of course. Even to the right group,
if somebody would tell me what that was!
Thanks
------------------------------------------------------
Ross M. Greenberg @ NYU ----> { allegra,ihnp4 }!cmcl2!acf4!greenber <----