Dr. Dobb's | The Real Deal | December 01, 1997

The Real Deal

December 01, 1997
URL:http://www.drdobbs.com/the-real-deal/184410352

Dr. Dobb's Journal December 1997: Editorial

Like the Energizer bunny, some good ideas keep going and going and going...Take the "card" user-interface metaphor, for example, which cropped up in the early 1980s, when researchers were looking for user-friendly (and, more often than not, hypertexted) ways of organizing and presenting information. With "Zoomracks," Paul Heckel was among the first to implement a card-based system that used Rolodex-like "racks" to hold "cards" that could be "inserted, removed, and moved into other slots in the same or different racks." (Heckel described the concept in his November 1985 DDJ article, "Zoomracks: Designing a New Software Metaphor.")

The idea behind Zoomracks was (and still is) a good one. So good, in fact, that others -- with or without knowledge of Zoomracks -- built similar card-based schemes of their own. Frank Halasz, for instance, designed "NoteCards" at Xerox PARC in 1986. Developed on Xerox Lisp machines using the InterLisp programming environment, NoteCards was a hypertext system based on cards, "fileboxes," and the links between them.

But it was Apple Computer that put cards on the mainstream table in 1987, bundling its hallmark "HyperCard" free-of-charge with every Macintosh. Although not originally designed as a hypertext system, Apple added linking, external commands, and the HyperTalk language, and HyperCard immediately gained a fanatical following.

When Heckel received a patent on the card metaphor, however, he turned a good idea into a controversial one. Further fanning the flames, he sent saber-rattling letters to Apple, Asymmetric, IBM, and others. A legal catfight ensued, and Apple ended up paying Heckel royalties on HyperCard sales.

Still, references to cards kept popping up, including Tim Berners-Lee's 1989 proposal for the World Wide Web. Although, admittedly, HyperCard wasn't among the hypertext systems he originally examined (Guide, Microcosm, Griff, KMS, and others were), Berners-Lee does acknowledge that the now-ubiquitous phrase "home page" links back to HyperCard's "home stack."

HyperCard competitors began to appear in the early '90s. "SuperCard," for instance, imported and converted HyperCard programs, included the SuperTalk scripting language, and let you create links between cards and other objects. A forthcoming SuperCard Runtime for Windows from Allegiant (http://www.allegiant.com/) promises to run Macintosh stacks on PC platforms.

The most recent incarnation of the card user interface is "Lifestreams," an Internet-based system that organizes information in terms of time-ordered streams, rather than desktop space. Developed by a Yale University team headed by David Gelernter, Lifestreams looks somewhat like a winning hand of Windows Solitaire, with cards bouncing across the screen. Beneath the cascading-card interface, however, is a virtual machine that's independent of the UI above and the database below. Each card is an individual document, with the most-recently created or received ones in the forefront. You browse the cards by moving the cursor along their edges, thereby triggering document summaries. Clicking on a document lets you view it or launch the application that created it.

Lifestreams includes a universal storage system, browser-like searching and filtering, and a data model that integrates information. Filters and agents organize, locate, summarize, and monitor incoming information. The system supports five basic operations. find lets you enter criteria for creating a "substream" (a set of documents extracted from the archive); squish compresses a substream in a content-dependent context; new creates a new card; clone makes a copy of an existing one; and transfer lets you move a card from one lifestream to another.

The goal of Lifestreams is to organize the electronic "bits of paper" that find their way to us, whether they be e-mail, images, web pages, or whatever. To this end, the fundamental problem Gelernter and crew are tackling is the inability of the 20-year-old desktop user-interface metaphor to alleviate the confusion that results from too much information. Apple fellow and UI guru Donald Norman concurs, stating flat out that "the desktop is dead." Norman went on to tell Wired magazine writer Steve Steinberg that "today, we have thousands of thousands of items -- more than can possibly fit on the screen." This is exactly the problem that Gelernter sees Lifestreams as solving.

Along with Eric Freeman, Gelernter has founded Mirror Worlds Technologies (http://www.mirrorworlds.com/), a company focusing on interfaces, architectures, and tools for network, information, and distributed computing. Although Lifestreams currently runs only on UNIX systems, Mirror Worlds (a name borrowed from Gelernter's 1992 book Mirror Worlds: Or the Day Software Puts the Universe in a Shoebox...How it Will Happen and What it Will Mean) plans on releasing a Java implementation before long.

You'd think an idea that's good enough to keep coming back might eventually catch on -- and Lifestreams might be the ticket that deals "cards" a winning hand. In any event, alternative user interfaces will continue to crop up as long as computers continue to create as many problems -- like burying us with information -- as they solve.

--Jonathan Erickson