1



Specifications for a MOO-based WWW server for work groups


Background of the Institute for Advanced Technology in the Humanities

The Institute for Advanced Technology in the Humanities was established in 
1992 to provide researchers in the arts and humanities with an opportunity to 
employ sophisticated technical support and advanced computer technology in the 
service of their scholarship.  The Institute encourages collaborative research 
proposals from humanists on the internet, and it can offer creative 
technologists a broad range of research opportunities.  The Institute is 
unique among humanities computing facilities for its combination of a broadly 
interdisciplinary charge with a projects-centered and research-oriented plan 
of work.  In addition, it is unusual for its success at bridging the gap 
between the two cultures of academe:  from its inception, the Institute has 
depended on the synergy of computer science and the humanities; its fellows 
meet regularly with computer science faculty and graduate students, and they 
confer and collaborate on a daily basis with the Institute's staff of computer 
professionals.

The principal goal of the Institute is to see that humanists have the tools, 
the training, and the support to make the most of computer technology.  In 
part, this entails a consolidation of existing facilities in order to produce 
an integrated environment for the use of networked information resources.  In 
addition, the Institute's support is often concentrated on converting 
information into electronic form and addressing issues of standards and 
formats that arise in the process of this conversion.  Finally, where existing 
software does not meet the needs of scholars in the humanities, the 
Institute's technical staff will work with the Institute's Fellows to produce 
broadly useful software tools.

The Institute also addresses itself, wherever possible, to topics and 
challenges of general importance for electronic scholarship across the 
disciplines.  These include (but are by no means limited to) testing methods 
for collaborative networked research, peer-review, and editing, investigating 
the possibilities for networked hypermedia archives, developing practical 
models for cost recovery in networked scholarly publishing, exploring the 
creative potential of the networks, negotiating cooperation with the private 
sector and with government, studying the implications of the digital library, 
and addressing the impact of electronic media on intellectual property, with 
regard to both the author's right to credit and the user's right to access.

Recent research projects at the Institute include:

	Ed Ayers, "Valley of the Shadow," a Civil War archive 
	Jerome McGann, "Rossetti Archive" 
	Ellen Contini-Morava, "Noun Classification in Swahili"
	John J. Dobbins, "The Pompeii Forum Project"
	Hoyt Duggan, "Piers Plowman"
	Mike Gorman, "Bell's Path to the Telephone"
	Gary A. Anderson, "The Life of Adam and Eve:  The Biblical Story in 
		Judaism and Christianity"
	Ken Schwartz, "Charlottesville Urban Design and Affordable Housing"
	Judith Schwartz, "The Hierarchical Audio Construction Kit:  Composition 
		and Design"	
	Michael Stern, "Visions for a Sustainable City:  Ownings Mills, MD"
	Elizabeth Crocker, Sprint Fellow, "Identity in the Age of Mechanical 
		Reproduction:  Technologies of Assimilation in the American City, 
		1880-1940"

These fellows have employed and created network technologies for the 
advancement of scholarship in the disciplines of literature studies, history, 
religious studies, architecture, urban planning, and anthropology.

History of MOOs

MOOs came into existence ten years after the invention of UNIX, when Pavel 
Curtis and others working at Xerox PARC invented MOOcode.  MOOs belong to a 
class of programs known as MUDs (Multi- User Dungeons), which are network-
accessible, multi-participant virtual reality programs whose user interfaces 
are entirely textual.  In the MOO, really just a large database, the world is 
made up of objects:  objects always have a particular location within the 
database and the landscapes of the MOO, and objects are in turn composed of 
properties and verbs.  The "OO" in MOO stands for "object-oriented."  Like 
UNIX, MOO functions as a command interpreter and programming environment.  
Programmers can easily have access to the individual building blocks of the 
MOO and can use the tools within the MOO to alter or add to those building 
blocks.  MOOcode and UNIX are similar in that both are hierarchical, multi-
user, time-sharing environments for the creation, storage and retrieval of 
information.  

This family of software, which was first seen and used primarily as a vehicle 
for games, now shows great promise in many arenas as a tool for network 
communication and collaboration over the Internet.  MOOs function with a basic 
hierarchy much like that of UNIX.  MOO-characters are divided into three 
ascending classes:  player, programmer, and wizard, with each class 
representing a higher level of programming control.

In addition to their original incarnation as textually mediated communication 
and virtual reality programs,  MOOs are now increasingly used as WWW servers.  
MOO-based WWW servers do not allow for real-time communication between users 
but do provide some editing capabilities.  The technology is currently limited 
such that users can only edit the objects they create and own; that is, users 
can add to the database (for instance, in the form of comments), but they 
cannot change the overall parameters of the MOO or redesign pages already 
posted on the Web.  And, in order for the users of the MOO to create pages, 
they must be able to program in MOOcode. 

The potential for MOOs to serve as tools of genuine collaboration, then, is 
currently limited by the fact that the software does not permit a hierarchical 
distribution of authoring permissions for non-expert users, as would be 
beneficial for many research, business, publishing, and classroom projects.  
What is missing, in short, is a user-friendly administrative interface for 
assigning and controlling read-write permissions on collaborative, MOO-based 
Web documents.  

Proposal

What we propose is a significant refinement of MOO-based Web servers.  With 
the AT&T Foundation grant, we will create four new types of MOO-interfaces to 
aid in the creation of collaborative Web pages.  These four types are
	
	1)  an administrative interface that can assign different read and/or 	  
	write permissions to different users or work groups. 
	2)  a submissions interface that will allow for the easy submission of 
	pre-existing html and vrml documents to a networked database
	3)  an authoring interface that will give users the ability to edit 
	existing Web pages, depending on their level in the project hierarchy
	3)  a scrapbook interface that will enable users of the MOO to collect 
	their own excerpts of documents

The new administrative interface will serve the various needs of several 
classes of users working on a networked project  by assigning read and/or 
write permissions on particular documents or parts of documents appropriate to 
their role in the collaborative process (whether it be peer review, archive 
creation, proposal writing, etc.).  For example, in a peer review project that 
employs such a server, reviewers will be able to comment on an article but not 
alter the text of the article or see the comments of other reviewers; authors 
will be able to alter the text of their articles and read but not alter the 
comments of reviewers; editors will be able to read and alter the text of all 
comments and the article itself.  Users of the MOO will be assigned different 
permissions based on their being identified as members of particular groups.  
Documents will similarly be classed such that they are partially or entirely 
open for some groups, but not others.  Basic administrative forms for creating 
these different groups using standard WWW clients will also be provided.

This type of user-interface will serve to address the needs of non-expert 
administrators who desire stratified levels of authorial control, and it will 
be well suited to projects for which selective and highly articulated access 
is a significant asset.  Current Web servers do not have the ability to retain 
state information; users are simply allowed or denied access, and their 
preferences are not recorded.  The authoring and submissions interfaces this 
project will create will enable users who identify themselves within the MOO 
hierarchy to take advantage of gradations of access and authoring permission.  

Overall, as the paradigm for Web use changes from browsing to one of 
authoring, as seems destined to happen, this tool will help make the newly-
embraced possibilities of the network easier to implement for the average 
user.

Rationale

Several recent and ongoing projects at the Institute have pointed to a strong 
demand for software of the kind described above.  Fellow Ed Ayers is in the 
process of creating an electronic hypermedia archive entitled "The Valley of 
the Shadow:  Living the Civil War in Pennsylvania and Virginia"  which 
compares how the cities of Staunton, Virginia, and Chambersburg, Pennsylvania 
weathered the Civil War.  Ayers and networked colleagues in different 
geographic locations have used email extensively for communication purposes 
but, because of the limitations of the MOO software for non-expert users, they 
have not employed MOOs for the editing itself.  Ayers also desires a scrapbook 
interface for his project, so that users will be able to excerpt personalized 
archives from the already posted documents.  

Likewise, fellows Joseph Viscomi at the Institute, Morris Eaves in Rochester, 
and Robert Essick at the University of California, Riverside will be working 
in 1995 on a hypermedia archive based on the illuminated books of William 
Blake with support from the Getty Grant Fund and the Institute.  A MOO-based 
Web server that could assign different writing permissions to users working at 
different levels of the compilation process would be an enormous benefit to an 
enterprise of this kind and scope.  With researchers working in different 
parts of the country and the world, a MOO-based Web server that could address 
the communication and editing aspects of the collaborative process 
simultaneously would greatly enhance the cooperative creation of a networked 
document.    

The utility of this software is not limited solely to research projects.  MOO-
based Web servers also have great potential as collaborative teaching tools.  
This past academic semester, University of Virginia professor Michael Stern's 
graduate landscape architecture class (LASR 702) established a network 
collaboration between architecture students at the University of Virginia in 
Charlottesville and at Catholic University in Washington, D.C. to create a 
design-consultation hypertext document, "Visions of a Sustainable City."  The 
project employed a forms-based interface and cgi-scripts written in perl to 
facilitate the collaborative creation of Web pages.  This interface, which 
proved somewhat difficult to work with from a technical standpoint, was 
created and tailored specifically for this project and could not easily be 
generalized to other uses.  The MOO-based Web server the Institute is 
proposing to create will not only meet the same requirements but enhance the 
overall functionality of MOO-based collaborative authoring and eliminate the 
need for any project-specific programming of this kind.  This software would 
also be useful for classes such as John Unsworth's undergraduate seminar "The 
Practice and Theory of Hypertext" (ENSP 482), for which students collaborated 
on various hypertextual, creative documents.

A MOO-based server with the access-defining capabilities described above would 
additionally be of immediate use in the peer review of the Institute's 
electronic journal Postmodern Culture.

The idea and perceived need for this kind of software grows out of David Blair 
and Tom Meyer's work on the hypermedia document "Waxweb," an electronic film 
with an authoring interface which allows users to collaboratively add to the 
story.  "Waxweb" is at present the longest hypermedia narrative document on 
the Internet, and it is accessible to both WWW (Mosaic) and MOO users.  The 
creators of Waxweb refer to it as a "hypermedia" document because it 
incorporates hypertext, visual, audio, video, and virtual reality objects, the 
latter using virtual reality modeling language (vrml).  Currently, "Waxweb" 
allows users to add comments and see others' comments, but the hardcoding 
denies them the ability to edit existing pages, something that even the 
creators of "Waxweb" can do only in MOOcode.  The creators of "Waxweb" 
identified the need for a user-friendly administrative interface that could 
facilitate a hierarchy of programming permissions.

Description

The objective of the project will be to create working piece of software in 6 
months for free distribution to all users.  Three part-time technical staff 
members will be hired (2 at the University of Virginia and 1 at Brown 
University) to assist with the actual programming with associate fellow of the 
Institute David Blair serving as a consultant.

Evaluation

This project will be deemed a success by the Institute when the participants 
succeed in developing a MOO-based Web server that can provide selective access 
with the high degree of sophistication described above.  This project will 
demonstrate the viability of the WWW as a collaborative authoring and teaching 
medium.  Once this software exists, its potential applications in the 
classroom are numerous, whether for networked collaborations between students 
in different geographic locations, for faculty research projects, or for 
electronic, scholarly publishing.  With the strong push in American 
universities to promote hypertextual awareness and the educational 
applications of Web, this project comes as a timely aid in making the Web more 
accommodating to collaborative projects, particularly those involving non-
expert users.

Budget

Much of the work to develop this server has already been done by Tom Meyer and 
others at Brown University, working in concert with the Institute for Advanced 
Technology in the Humanities and under the direction of David Blair, a 
networked associate fellow of the Institute.  The budget for this project 
includes funds for a student working at Brown under the direction of Tom 
Meyer, staff working at the Institute and elsewhere under the supervision of 
John Unsworth, Director of the Institute for Advanced Technology in the 
Humanities, and consulting from David Blair.  The technical director for this 
project would be Tom Meyer, and alpha testing would be done with the 
electronic journal Postmodern Culture, co-edited by John Unsworth.  

The project should be completed, to a testable stage of development, within 
six months.

1 tech. staff (Brown)    	@ 	$24/hr, 	20 hrs/wk, 	25 wks:$12,000
2 tech. staff (IATH)    	@	$12/hr,	20 hrs/wk,	25 wks:$12,000
Consulting (David Blair) 	@	$25/hr,	10 hrs/wk,	25 wks:$6,250
Subtotal:									       $30,250

Equipment:		RS6000 RAM						 $4,750
			because MOOs are RAM-intensive
			[50 mb dm = 170 mb RAM]
Total:										            
$35,000


AT&T Rationale

This project addresses AT&T's interest in networked computing.  Interest in 
the WWW is currently very high (WWW recently surpassed email as the largest 
portion of internet traffic), and the MOO software we are proposing would help 
make the Web accessible to a wider ranger of users as a medium for 
collaborative authoring, and not just as a browser for static information.  
Additionally, it will benefit curriculum development, as educators will be 
able to use it to generate hierarchies for the cooperative authoring of Web 
documents.  Most of the projects at IATH that have shown a need for this kind 
of software are by nature interdisciplinary, and the interest in this tool 
grows directly out of users' concerns and their experience with existing Web 
technologies.