kovar@biostat.harvard.edu (David C. Kovar) (04/29/91)
Someone at Harvard brought this to my attention. Offhand, it seems like
a good way to beat the Internet to death in one move. Like flushing all
the toilets at DEC and watching the plumbing explode. I don't know
exactly how they implemented it, but the parallel approach and the decending
search of a tree approach sounds like they're going to start at the top of
the domain name servers and work they're way down, hitting each server
below a node with a new request. Anyone want to guess at how many nodes
one request would touch, and how much traffic would be generated?
There also might be a legal side to this, but that's a lot more fuzzy.
I suspect that the research networks would not take kindly to someone
making money by loading down their links, but then again, these guys
are just selling a tool, not a network based service.
-David Kovar
------- Forwarded Message
Subject: NetFind, whitepages service for the Internet
From: xcaret@csn.org (Xcaret Research)
Newsgroups: comp.newprod
Organization: Colorado SuperNet Inc.
NetFind, from Xcaret Research, Inc., is a whitepages service for the
Internet. Given a person's name and location or organization, NetFind
returns their Internet electronic mail (email) address. NetFind
differs from other whitepages facilities by offering fast and up to
date service from your own machine.
NetFind runs from the UNIX command line. To find John Smith at
Bigstate University, you just type:
% netfind smith bigstate
NetFind returns the email address. Its that simple.
NetFind operates on a Sun3 or Sun4 computer running SunOS 4.0 or
later with access to the Internet. Uses only 4MB of disk space.
Since NetFind runs from the command line, you can use NetFind from
any remote terminal, Mac or PC that accesses a Sun.
NetFind uses a unique method to actively search the Internet for
your target. It does not attempt to keep a database of users across
the Internet; that would be quite large and constantly out of date.
Instead, NetFind uses the natural database of the Internet itself: it
sends multiple parallel requests across the Internet to machines
where it suspects the target may reside. NetFind queries these
machines for the name you provide as well as for leads on other
machines nearby. NetFind in turn searches these discovered
machines and so on down the hierarchy until the address is found or
there are no more machines to search. The whole process is
surprisingly fast, because NetFind sends searches out in parallel.
NetFind uses a reference file for domains and a location index. They
take up about 3MB. The first file is a partial list of Internet
domains. NetFind uses this list to prime its search. The list is by
no means all of the computers it can search. Rather, it will use the
information obtained from querying those machines to further its
search.
The second file is an index of location words. When you say:
% netfind smith bigstate
NetFind looks in this index for "bigstate". Then it resolves the
information it sees with specific machines in the domain file and
begins its search. So, the whole process goes from location word to
index to domain file and then in parallel searches throughout the
Internet.
The name can be either a last name or a username. The location
words are generally geographical or organizational (i.e., switzerland
or sun) and can be combined to narrow the search
(i.e., netfind smith bigstate university ).
NetFind is sold with a single user license for $159. Please call for
educational discounts and site licensing. You may order by phone, e-
mail or postal mail (include P.O. number, check or credit card).
Please specify your domain name, Sun3 or Sun4, and the format
(electronic, floppy, 60MB, or 150MB tape). We don't charge for
floppies, but cartridge tapes are an additional $30 to cover cost.
Xcaret Research, Inc.
2060 Broadway, Suite 320
Boulder, CO 80302
(800) 736-1285
netfind@xcaret.com
------- End of Forwarded Message
--
-David C. Kovar
Consultant ARPA: kovar@biostat.harvard.edu
Eclectic Associates BITNET: corwin@harvarda.bitnet
Ma Bell: 617-643-3373 MacNET: DKovar
"It is easier to get forgiveness than permission."droms@SOL.BUCKNELL.EDU (Ralph E. Droms) (04/30/91)
Xcaret does not depend solely on a descending tree search of DNS; it
uses a collection of heuristics to determine where to start searching
and then may elect to extend its search through DNS.
The work on which Xcaret is based is described in:
Michael Schwartz and Panagiotis G. Tsirigotis, "Experience with a
Semantically Cognizant Internet White Pages Directory Tool", Journal
of Internetworking: Research and Experience 2, 1 (March, 1991).
- Ralph Droms Computer Science Department
droms@bucknell.edu 323 Dana Engineering
Bucknell University
(717) 524-1145 Lewisburg, PA 17837heimlich@watson.ibm.com (Steve Heimlich) (04/30/91)
A description of this product seems to appear in the March 91 issue of Internetworking Research and Experience, pp 23-50. The article is "Experience with a Semantically Cognizant Internet White Pages Directory Tool," Michael F. Schwartz and Panagiotis G. Tsirigotis (U. of Colorado at Boulder). I believe this paper describes what was announced. Steve