... MaxSearch would support a variety of interfaces, including a native or command-driven style, a more prompted textual mode, and a windowed, graphical environment.
It all started with Maxwell Online. Following Maxwell's acquisition of ORBIT Search Service, then of BRS, senior members of the Maxwell organization went on record last year as saying that it would be just as easy to build a new system from the ground up as to merge the two existing services. For the first time in twenty years, there appeared a chance to break free of some old assumptions and constraints, and to take a fresh look at what an online service should be.
Searchers lost no time in taking advantage of the opportunity Maxwell had presented. At its 1989 retreat, the Southern California Online Users Group (SCOUG) addressed the question, "What should the ideal online system for the 1990s look and feel like?" Many of their comments and suggestions were summarized in the first part of this article (ONLINE, March 1990). Highlights include:
"Something better than Boolean:" a
natural language retrieval process
built on relevance
Less emphasis on individual data
- bases, and more on getting the data
regardless of where in the system it
* Graphical interfaces, image support,
hypertext and hypermedia
Cost schedules based on information
retrieved, rather than on connect
This "searcher wish list" encompassed proposals ranging from the practical to the utopian, and raised a couple of key issues about the shape of online information retrieval in the 1990s:
To what degree should enhancements reside locally, i.e., in the user's microcomputer versus on the online service's mainframe? If more powerful features are locally-resident, how do we handle the proliferation of front end software packages, each with its learning curve and memory requirements?
Are enhancements being planned with information specialists in mind, or with an eye toward capturing a larger base of end-users? If the latter, what are the implications for the development of online services during the 90s?
I asked representatives of most of the major U.S. online services, as well as two European hosts, to address these issues. I questioned them about their short and long-term response to user concerns, and on a broader level, about the directions in which they thought their services, and the online industry in general, would be moving in the next decade.
On the whole, my respondents were straightforward, insightful, and generous with their time. I will summarize our conversations, then compare the issues and goals being addressed by the databanks with the concerns articulated by users and described in the first part of this article. Finally, I will discuss, very briefly, some of the technology- and market-related issues that will affect the shape and direction of online services in the '90s.
MAXWELL GETS IT ROLLING
Maxwell Online, faced with the problem of integrating ORBIT and BRS search protocols, will continue to support both search engines, while enhancing each service with appropriate new features. The much-anticipated brand new system," known internally as MaxSearch 2000, is envisioned as a seamless gateway between the two services, allowing the user to search BRS using ORBIT commands, and vice-versa. According to Joseph Paulson, Vice President of Software Development for Maxwell Online, MaxSearch would support a variety of interfaces, including a native or command-driven style, a more prompted textual mode, and a windowed, graphical environment. Much of its power would reside in the newly-designed mainframe system, but many features would be workstation-based.
Paulson identified the following specific system enhancements as priorities for the 1990s:
Enhanced searchability via built-in equivalencies, e.g., singular-to-plural, spelling variants, synonyms. …