At 10:30 PM 4/22/96, Vadim Antonov sez:

The idea of using "randomizing" sub-domain is to get people used
to the idea of randomization in domain names. Then it can be
moved into protocol or implementations. It also fixes the name
collision and transparent introduction of multiple registries.

Wasn't that the purpose of the naming scheme? Since local
laws already account for no two entities having the same name within the
same city, it would seem to solve the problem. Just wondering why we'd
reinvent the wheel with randomizing domain names unless it's to serve a
vendetta against the overregistration of COMs.

Then again, how long did the WELL go by before they started
using on all their official documents? Marketing departments want
to play it safe. If some percentage of their sales leads can't remember, that's lost business; that's all they see, the dollar signs
getting smaller.