     WAIS - The Scholar's Apprentice
     ===============================

One of the antecedents for today's scholar and scientist was the sorcerer of 
times long past. Unlike the workplace of today's researchers (wink wink), the 
studies of sorcerers and alchemists were generally untidy places with plenty 
of eye-of-newt and horn- of-toad lying about. To keep some order, apprentices 
were kept to clean the vessels, stack the vellum and get more eye-of-newt, 
tedious tasks that have the modern equivalent in rummaging through electronic 
library catalogs and browsing gopher menus. Enterprising sorcerer apprentices 
would occasionally mobilize legions of brooms to accomplish their tasks 
automatically.

Enterprising electronic apprentices of today have come up with WAIS - Wide 
Area Information Service - to accomplish the task of searching many electronic 
resources automatically. WAIS is a modern, electronic demon which will keep a 
list of electronic sources in its memory, accept a key-word search, connect 
and search each electronic resource automatically, and, finally, display the 
results ordered according to most hits found.

Today, WAIS is barely in its infancy. International standards are working to 
extend the functionality of WAIS to allow more complex boolean searches, to 
allow more sophisticated strategies of finding similar documents, to allow 
multiple document types and also to allow charge-back arrangements.

It is thought that WAIS - or something that works like it - will be the way 
information both public and private will be disseminated in the Global 
Electronic Village. It is quite easy to contemplate that a tireless, 
electronic, artificially intelligent assistant would be running on your 
computer. This assistant will have digested every document or subject in which 
you have expressed interest in the past, and its only satisfaction is to find 
more things that might be of interest to you and present them several times a 
day.

========================
The Requirements of WAIS
========================

WAIS runs on multiple platforms, PC's, MAC's, Workstations, X- Windows and on 
JHUNIX. The only real requirement is that the host machine have and ethernet 
card and be connected to the Internet. There are quite a few anonymous ftp 
sites around the world that have WAIS programs for various platforms. In order 
for WAIS to work, everybody has to use it. That is why the programs are 
generally free. Once a user community has been created, the designers of WAIS 
hope to sell commercial add-ons that make WAIS perform better. Companies that 
sell information also hope to give away plenty of razors to sell plenty of 
blades.

=========================
Getting Started With WAIS
=========================

As an example we will pick PCWAIS, a WAIS program that runs on IBM compatible. 
PCWAIS requires the Clarckson packet driver for your ethernet card. PCWAIS can 
be downloaded from INFO.UNC.EDU in binary mode. You will get a file called 
PCWAIS.ZIP. When you unzip the file you get roughly 500k of program and the 
source code. To free up space you can delete all the files except the ones 
listed below.

PCWAIS.EXE - the executable file

PCWAIS.TYP - contains currently implemented WAIS document types

WATTCP.CFG - the configuration file for the network packet library WATT/TCP

DIRECTOR.SRC - WAIS sourcefile of the Directory of Servers maintained at 
quake.think.com

Before you start PCWAIS, you need to fill in a few items in the WATTCP.CFG 
file. Using your favorite editor, enter the correct information into the file. 
The fields my_ip, netmask, nameserver, and gateway each require an ip number. 
The field domainslist needs your correct domain inside of double quotes.

Below is an example of a typical WATTCP.CFG file for the JHU domain.

my_ip=128.220.2.10

netmask=255.255.0.0

nameserver=128.220.2.7

gateway=128.220.2.1

domainslist="jhu.edu"

The documentation is sparse; the authors stress two points in their readme file:

"If you are not sure of any of these values, ask your local guru -- incorrect 
numbers may appear to be working for you while actually wreaking havoc for 
other users on the network."

"PCwais is quite easy to use. Most of its features can be

understood simply by trial and (not too much) error."

===========================
THE PRINCIPLES OF OPERATION
===========================

A WAIS query has a certain structure that must be observed. First you need 
"source files." Source files contain all the information that your demon needs 
to connect and search a specific electronic resource.

Below is an example of a source files that contains patents:

Headline: patent-sampler.src S:1000, L:17

(:source

:version 3

:ip-address "131.239.2.100"

:ip-name "cmns-moon.think.com"

:tcp-port 210

:database-name "PTNT"

:cost 0.00

:cost-unit :free

:maintainer "bug-public@think.com"

:subjects "general patents"

:description

"Connection Machine WAIS server.

About 2 weeks of Patent applications (18MBytes) from the US Patent Office.

"

)

Source files can be gotten automatically; however, it is important to become 
familiar with sources in you field. There is generally less wasted motion when 
a researcher is intimately familiar with the sources. There is generally less 
chaff in the hits.

Each source file you want to include in your library of electronic sources 
will have the extension .src and will reside in your WAIS directory. At 
present, WAIS sources are strongest in the technological disciplines. It is 
easiest to find reviews of specific computer programs and discussions of 
electronic methodologies since many on-line discussions are indexed. Economic 
data and discussions of trends are also strongly represented. There is a 
smattering of sources in the humanities: the Bryn Mawr Classical and the Bryn 
Mawr Medieval Review are two new entrants that will index primarily book 
reviews.

A good strategy to get a sense of the myriad sources is the use the one source 
file that comes with PCWAIS, director.src to find others. The next few 
paragraphs will be a tutorial to help you get started.

================
STARTING A QUERY
================

When you launch PCWAIS the program presents an empty screen with a menu bar at 
the top and some shortcuts on the bottom. By clicking on the "Question" menu 
or pressing [F7], you can start a new question. Notice that questions can be 
saved to be searched or amended at a future time when new sources may be 
available.

When you initiate a new search you see the NEW QUESTION window. This window 
has two secondary windows on top and a "Results" window on the bottom. In the 
center of the screen there is a text box that allows you to enter the text 
string you would like to search. The first top left window contains a list of 
all the .src files in our wais directory. Notice that we only have one at 
present, the "Directory" that comes with the program. Our first step will be 
to get all the .src files that are listed with the "Directory of Servers" at 
think.com.

By clicking on the one entry in the top-left window and clicking on the >> Add 
button, we can move this .src file from the list of all our sources a subset 
that will be used in this particular query. Next we click on the text box and 
type in the letters 'src' - that means we want think.com to bring up a list of 
ALL the .src files it carries. To start the query, click on Run. The program 
will flash several messages on the bottom of the screen to tell you that it is 
connecting and searching. When it is finished, the 'Results' box will fill up 
with the 'hits' - in this case, a list of all the source files available.

If you want to see one, scroll down the window, click on the one you want and 
click 'Retrieve.' Up pops a new window with the text of the scr file. If you 
want to consult the ANU economic papers in the future, you can save that file 
to your WAIS directory and the next time you start WAIS, it will show up in 
the top-left window.

================================
COLLECTING and SEARCHING SOURCES
================================

Let us assume we have collected several sources. They all appear in the top-
left window. This time we want to search the 'Science Fiction' reviews for any 
mention of Robert Heinlein:

1. We delete the 'directory' in the 'Sources' window by clicking >> Del

2. We add the sf-reviews

3. We type 'heinlein' into the text box in the center

4. We click Run and wait for the results to show up.

5. We browse the results by clicking 'Retrieve' and scrolling through the 
document.

6. Any documents we want to keep can be saved to our disk.

==================
OTHER USES OF WAIS
==================

In keeping with its role as electronic demon, WAIS has other uses. For 
example, it is possible to set up a system where WAIS will automatically index 
your incoming and outgoing e-mail. A key-word search is a easy method of 
locating month-old correspondence in moments.

=======
SUMMARY
=======

The principles of WAIS are determined by the nature of electronic information. 
Gone are the days when data would reside on one server. Today, data may reside 
on twenty different servers. By managing collections of src files, it is 
possible to search ten geographically dispersed servers with one query. You 
WAIS demon will connect to each one and download the results to your machine. 
One of the most promising feature of advanced WAIS programs is the ability of 
the program to find documents that are similar to one which you found helpful. 
Thus you don't have to devise a clever key-word search, you simply say to you 
electronic apprentice - go find me some more of the eye-of-newt you got last 
week.

If you have a network connection and would like to get WAIS working on your 
system. please call Peter Batke at 8096. 
