Hal Finney:

I think it would be better to only provide anonymity when asked.

Johan Helsingius:

It all depends on the intended target audience.

I have to agree with Johan here, and with the way all (?) the Aserver creators/administrators have chosen to go. Consider the following scenarios, assuming that Person-X does not know anything about the particulars of the Aserver(s) he/she is routed through when making a direct *reply* to an anonymous message. A) Person-X doesn't care if he/she is aliased when he/she replies, but he/she is aliased anyway. consequence: Not much, the message still gets through (as would a re-reply, so if the Person-X *wants* to make his/her identity known later he/she can always state it in a message body). B) Person-X *wants* to be aliased in his/her reply, but *isn't* because the Aserver doesn't do it automatically, and person-X isn't aware that such a "switch" needs to be thrown. consequence: Potentially disasterous to person-X! I submit that automatic aliasing, by default, is consistent with the very purpose of Aservers, more exactly their intended, legitimate, uses! This doesn't mean that Aliasing software shouldn't contain some provision for turning ON a switch that passes you through un-aliased, but this switch should be for users who KNOW the server and how to modify its default behavior. I wouldn't object to such a switch, but personally I don't see much use for it either. Once I knew I wanted to reveal myself to someone, I could just tell him/her in a message body. If they want to reveal themselves to me, they can do likewise, and then we can address each other's machines directly, bypassing the Aserver(s). Now if these same creators/administrators would only *turn off* the automatic (default) message saying: "An alias [ALIAS####] has been created for you on Aserver@somewhere.in.the.world" the privacy of what seems to be a growing, potential, aliasing network (Anetwork) would be significantly enhanced. Reversing the default here would be consistent with the "most conservative assumption" approach otherwise already taken with respect to auto-alias. matthew rapaport Philosopher/Programmer At Large KD6KVH mjr@netcom.com 70371.255@compuserve.com