>Path: funic!fuug!mcsun!Germany.EU.net!news.netmbx.de!mailgzrz.TU-Berlin.DE!math.fu-berlin.de!fauern!ira.uka.de!sol.ctr.columbia.edu!The-Star.honeywell.com!umn.edu!news2.cis.umn.edu!gopher-news-daemon at boombox.micro.umn.edu>Date: Wed, 24 Feb 93 11:00:15 CST
>Message-ID: <9302241700.AA19880 at boombox.micro.umn.edu>
>From: mpm at boombox.micro.umn.edu ("Mark P. McCahill" )
>Reply-To: "Mark P. McCahill" <mpm at boombox.micro.umn.edu>
>Original-To: dsm at ibiza.cc.columbia.edu>Original-Cc: gopher-news at boombox.micro.umn.edu>Subject: gopher licensing
>Sender: news at news2.cis.umn.edu>Approved: comp.infosystems.gopher at news.cis.umn.edu>Lines: 226
In message <9302241503.AA05488 at ibiza.cc.columbia.edu> David Millman writes:
>> The following article may stir rumors that you (or U Minn) may intend
> to start charging or licensing the gopher software or protocol. (It
> doesn't say so explicitly--actually it only mentions the 'cost'
> attribute in gopher+) But the context is a little scary (especially
> given the direction that Brewster is taking).
>> Before you are poorly misquoted, can you clarify any such intentions?
This is going out to a fairly wide audience so that we deal with the issue
once rather than a bunch of times.
There is some information that will only be made available online
if the information producers can charge for access. While lots of information
wants to be free (look at all the stuff currently available via anonymous ftp
and gopher), some information isn't going to be available unless there are some
mechanisms for charging. We want gopher to be a tool that allows you get at
both flavors of information... and we have this funny feeling that if our
software is being used for commercially, we ought to get something in return.
We have been able to justify making gopher freely available to the higher
education community based on the idea that it makes more information available
to us (and everyone else) on the internet. However, when someone starts making
money from our work, it makes sence for the University to get a piece of the
action. In fact, we have already done some deals already with the
commercial world for use of our software, and hope to do more deals in the
One item that was not covered in much detail in the article (because of space
constraints) is an authentication method we built over the top of gopher+ that
allows for user authentication WITHOUT sending the user's password over
the network. This scheme (called ADMIT1) is implimented in the TurboGopher
beta-test client for the Mac we are distributing and will be available for
servers and other clients in the very near future. Basically, the client and
server negotiate to prove that both the client and server know the user's
password, and this negotiation generates a ticket good for one access to an
item on the server... so a stolen ticket doesn't work. The client can generate
additional valid tickets without re-negotiating with the server. We are also
looking at supporting public key authentication. So... with authentication
methods available for restricted access servers, there will soon be for-fee
information available to compliment the free stuff.
> We are starting to rely quite heavily on your software, for which we
> and a lot of other people owe you big thanks. But we also need to
> plan for any potential costs--other than those imposed by individual
> servers--for using the software or, especially, the protocol.
> david millman
> coord of r&d, academic info systems, columbia u
Our plan is to continue to make gopher freely available to the education
community because this gets more free information available online for us
and everyone else on the internet. In the case of commercial use of our
software we are very interested in doing licensing deals because this gets
us the resources to do more development and support.
So... we are looking at expanding gopher use beyond the current higher
education community, and the best way we can see to get the resources to
do that is to get a revenue stream from the commercial world. We don't see
any reason to change what we are doing with gopher availability in higher
education because what we have been doing has worked well for everyone.
>>>>> -------------------------Cut Here--------------------------
>> Copyright 1993 Network World, Inc.
> Network World
>> February 15, 1993
>> SECTION: INDUSTRY UPDATE; Pg. 29
>> LENGTH: 1013 words
>> HEADLINE: Internet retrieval tools go on market;
> Gopher and WAIS to go commercial this month; other information retrieval
> products gaining ground.
>> BYLINE: By Ellen Messmer, Senior Correspondent
>> DATELINE: WASHINGTON, D.C.
> Three types of search-and-retrieval software tools currently available
> free on the Internet are being readied as commercial products.
>> In just over a year, Thinking Machines Corp.'s Wide Area Information
> Server (WAIS) and the University of Minnesota's Gopher software have been
> loaded onto servers at nearly 1,000 sites on the Internet, giving users the
> means to locate and retrieve information through text-based queries.
>> But even as the developers of WAIS and Gopher mull over plans to set up
> commercial ventures this month, a number of other products based on a third
> information retrieval method are gaining ground.
>> The library community is rallying around the 1992 ANSI standard Z39.50, an
> application-layer protocol for computer-to-computer information retrieval.
> The University of California,Pennsylvania State University and other colleges
> have set up a Z39.50 test bed on the Internet with AT&T, Notis, Inc., Digital
> Equipment Corp., Data Research Associates, Inc. and other vendors to develop
> interoperable products.
>> The first wave of Z39.50 products, including one from the On-Line Computer
> Library Center, Inc., hit the market just last month.
>> "Z39.50, like many such standards, is defined in the Open Systems
> Interconnection model," said Clifford Lynch, director of library automation
> in the office of the president at the University of California. But because
> Transmission Control Protocol/Internet Protocol is used far more than OSI on
> the Internet, Z39.50 was adapted to run over TCP/IP for the Internet
> prototype, he said.
>> While specific search and indexing features of Gopher, WAIS and Z39.50
> differ, all three technologies present users with a way to send out a query
> on a subject in order to find and retrieve information across a network.
> However, WAIS can be reached through Gopher, but Gopher cannot be reached
> through WAIS.
>> WAIS, in fact, uses the 1988 version of Z39.50 as its starting point, but
> is not compatible with the 1992 version of that standard. WAIS allows users
> on different computer platforms to find and access all types of information,
> including text, image or stored digital voice.
>> The client program can run on an end-user device or on a shared machine
> used as an access device, as is often the case on the Internet. The server
> does the indexing and retrieval of documents. All files or databases to be
> queried have to be indexed.
>> WAIS, Gopher and 1992 Z39.50 all eliminate the need to know the wide
> variety of commands to obtain files since the resource discovery tools can
> intuitively start a File Transfer Protocol session, or other access means,
> on the user's command to retrieve the data.
>> Thinking Machines, which developed WAIS, offers a version that runs on its
> powerful parallel computer, the Connection Machine. But the company's
> decision to release WAIS as a public-domain "freeware" package has spawned
> versions of WAIS client software for about a dozen of the most widely used
> computer platforms.
>> Although use of WAIS has increased during its use on the Internet, it was
> always intended to be an integrated search system for corporate users, said
> Brewster Kahle, project leader on WAIS at Thinking Machines. Now its creators
> appear ready to make WAIS commercially available for a variety of platforms.
>> Kahle said he will launch a company called WAIS, Inc. in Menlo Park,
> Calif., later this month in order to market commercial versions of WAIS.
> Although he declined to offer more details, he said the continued
> availability of freeware versions would help publicize WAIS and give users
> a chance to experiment with it. But he pointed out that companies want
> commercially available products with customer support.
>> The University of Minnesota, which holds the copyright to Gopher, allows
> free use of its client/ server software as long as the information on the
> server is made available free of charge on the Internet.
>> Public-domain versions of Gopher have been created for virtually every
> computer platform, ranging from personal computers to mainframes. "Gopher is
> designed to be easy to write clients and servers for," said Mark McCahill,
> manager of distributed computing and the project leader on Gopher.
>> Once files and databases are prepared with Gopher's inverted text-search
> software so that key words can be located, documents can be accessed from
> Gopher servers. Michigan State University started with a Sun Microsystems,
> Inc. computer as the Gopher server, and this week the university is adding
> an IBM RISC System/6000 Model 350.
>> Rick Wiggins, manager of the computer laboratory, said students and
> faculty use Gopher to search for anything from scientific data to library
> information and course schedules.
>> Internet use of Gopher has skyrocketed, according to Joel Cooper,
> assistant director of network services at University of Notre Dame, pointing
> to statistics from the National Science Foundation (NSF) that said there has
> been a 4,000-fold increase in Gopher traffic since last year.
>> And Gopher, too, is going commercial. According to McCahill, the
> university this month is releasing Gopher Plus, the next version. Added
> features will include a billing mechanism to charge for database use, support
> for binary text file transfer and a security method based on private-key
>> The issue of gateways between WAIS, Gopher and Z39.50 is a significant one
> to users. Last fall, the NSF funded a newly formed organization called the
> Clearinghouse for Networked Information and Retrieval to track development of
> the technologies.
>> As part of its work, the clearinghouse, based in Research Triangle Park,
> N.C., this month will release a public-domain version of Z39.50, which will
> be backward compatible with WAIS.
>> "It's not a one-solution world," said Jim Fu