On Sun, 2 Oct 1994 12:47:25 EDT, Anthea Tillyer <[log in to unmask]> said:
>As a follow-up to my question about BYUVAX leaving BITNET, I would like to ask
>if there is a protocol (in the social, not computer, sense) for cases like
>this.
 
If I understand what you mean by "social," I'd have to say that it
depends greatly on the particular administrators of the particular
systems as to how "seamless" they make this process.  It all depends
on the level of commitment they have to their users.  Some choose (or
are forced) to take great pains to insure the transition is painless.
Others just "pull the plug."
 
>                                                              This is why
>I am asking if there is a "correct" way to do this
 
For BITNET nodes, Pete's already mentioned the LEAVING BITNET document.
But that describes the "correct" way in the "computer" sense.  I don't
know how deep it goes into the "social" and "moral" issues.
 
>       What obligation do list owners have in cases like this?
 
None.  Again, it all depends on the time, effort, and commitment you
want (or are able) to put into your list.
 
>                                                               Do we manually
>change each of those subscribers?
 
You can.  If you want.  And if you're able to figure out the new address.
That's not always very easy.  Actually, after the fact, it's pretty hard,
unless you have external knowledge of the host's particular situation.
 
>                                   Delete them all?
 
That's the other option.  If you're unable, or don't have time, to figure
out the new address, it's about all you can do.
 
>So, what is the "correct" way for sites, listowners, and users, to deal with
>address changes?
 
I've sort of been wondering that too, in a way...
 
If I could rephrase the question slightly:  Has LSoft given any
thought to implementing a BITEARN-NODES-independent way to do this
kind of thing?  There's nothing wrong with the :newnode approach,
except that it's not very useful for non-BITNET members.  But the
":newnode concept" is still a significant problem for the Internet as
a whole.  I'm not sure how feasible it would be for LSoft to tackle
the problem.  But it does seem that a solution would be valuable...
 
Bill
>Anthea Tillyer     City University of New York     [log in to unmask]