Until the introduction of Gribnif Software's NeoDesk 4, users of the Atari ST, TT and Falcon were left on the sidelines of the revolution in multitasking desktops for personal computers. NeoDesk 3, the previous version of Gribnif's alternative desktop, already supported multitasking, but only in the most basic way; as long as the computer was running Geneva, Gribnif's multitasking replacement for part of the Atari operating system, NeoDesk 3 was able to launch a new application while a currently running program remained active. But NeoDesk 3's own operations -- copying, deleting and formatting, for example -- did not multitask. And NeoDesk 3 did not have two other features that have made the best desktops on PCs so attractive -- a modern 3D look and feel, and a way of organizing and managing programs by groups.
NeoDesk 4 changes everything, and corrects an imbalance between the premiere desktop environment for the Atari and Microsoft's Windows. But in creating NeoDesk 4, Gribnif did not imitate Windows or fashion an Atari version of the Macintosh interface. NeoDesk 4 is more flexible and more intuitive than either of its popular counterparts. It is even able to perform some functions that Windows 3.1 and the Mac cannot ordinarily do.
And, almost as a tribute to the lean and efficient way Ataris have always operated, NeoDesk 4 occupies only as much memory as you want to yield over to the desktop. You can even run NeoDesk 4 comfortably on a 1-megabyte ST, although it will run more quickly and smoothly on systems with more memory.
A graphical user interface is nothing new. We like to point out with considerable pride that the ST arrived on our desks with a built-in graphical user interface long before most users of IBM-compatible personal computers had even heard of icons and windows. But the ST's GUI -- yes, it's pronounced "gooey" and not "Gee You Eye" -- followed the Mac's by a year. The ST's graphical environment was designed by Digital Research as almost a sideline in its efforts to create a GUI for PCs. This PC system, the Graphics Environment Manager, was stripped of much of its power and many of its features before it came to market, after a legal dispute between Digital Research and Microsoft, which was then developing its own graphical system called Windows. Microsoft complained that GEM was too much like Windows, and so Digital Research changed GEM. But the new owners of Atari Computer, casting about for an icon-and-windows interface for the exciting new 520ST, managed to convince Digital Research to leave the Atari version of GEM unsullied, and so the ST came to life with a GUI of its own, vaguely similar to the PC version of GEM. (That version disappeared from the marketplace in an avalanche of Windows.)
A graphical user interface does not have to use icons and windows, but that is how most of them have developed. It is easier to describe such a system as an "object-oriented" interface, because it allows the user to manipulate objects that perform actions -- actions that are, in most cases, analogous to what happens in "real life." Instead of typing commands onto the screen, the user instructs the computer to do any task by selecting an icon and doing something with it -- double-clicking on it or dragging it to another icon, for example. Rather than typing commands such as "XCOPY C:\BIN\FOO . /S A:" onto a blank screen, the user of an object-oriented interface merely deals with things -- objects of one kind or another, all represented by icons -- on the computer screen.
The way these icons were placed on the screens of the first experimental object-oriented interfaces in the late 1970s and early 1980s made the screens look a little like the desks in a typical office. They had icons for filing cabinets (disk drives in the computer system), for folders (directories on a disk), for a waste basket (the bit bucket, or the act of erasing a file), for a pile of notes, and for many other things. These early interfaces also had windows that opened up to show more icons or to hold running programs.
The metaphor of the computer-as-desktop is just one way of representing the way we deal with a personal computer. It may not be the best way, but it stuck. With the introduction of the Apple Macintosh in 1984 and the ST and Commodore Amiga a year later, the notion of a desktop on the screen began to gain popularity among users who had always viewed the standard PC command-line interface as crude and inadequate. In the IBM-compatible area, Microsoft's Windows took a halting step in that direction in its first two versions and then went a lot further in Windows 3.1 and in the latest version, Windows 95. IBM, creator of a hybrid interface called OS/2 1.0, also adopted the same sort of desktop in OS/2 2.0, 2.1 and OS/2 Warp and, at the same time, the GeoWorks company invented its own Mac-like interface for PCs -- one that actually surpasses the Macintosh in a dozen ways.
By the mid-1990s, "object-oriented" computer desktop interfaces had become, at last, the standard way of working with a personal computer.
Until recently, the ST's desktop interface, although easy to use, has been the weakest of all these systems. Although it uses icons and windows, the ST's GEM did not gain many of the full functions expected in an object-oriented interface until the release of version 2.05 of the ST's built-in operating system, known as TOS (for "The Operating System"). This was followed by version 2.06 when Atari added support for 1.44-megabyte floppy drives. A similar TOS-based GEM, which appeared as TOS 3.01 and was incremented to version 3.06, is built into the TT, and a newer TOS, version 4.xx, has been engineered into the latest Atari, the Falcon030.
But the GEM desktop built into the latest versions of TOS, while superior to the original version, lacks all the advanced features of the ranking monarch of alternative desktops, NeoDesk 4. Among the advantages of NeoDesk 4 are these features: