Home | Articles | Forum | Glossary | Books

The eyes and ears of the world.

If society is an organism, and if information is the lifeblood of the organism, then by comparison, its sensory systems must be viewed as primitive and disjointed. The effect is that of a beast that can in some limited ways see, hear, taste, smell and touch, but that is unable to transform those raw sensory stimuli into understanding, decision, and action.

A new strategic initiative, spearheaded by Microsoft but brought forth in conjunction with more than 60 leading vendors in most areas of technological interest, promises to link the eyes and ears of this organism with the rest of the nervous system. In so doing, it will endow it with the intelligence needed to evolve into a higher form of life, the type required for survival in the 21st Century. Table 1 summarizes some the major and many of the minor players whose support Microsoft has enlisted for MAW, or Microsoft at Work.

This initiative promises to enable communications among PCs, fax machines, fax boards, printers, copiers, telephones, and handheld computers. In this scenario, all those devices would be networkable and would contain lots of intelligence. All would contain some form of the Windows GUI (Graphical User Interface) to enable features that in many cases already exist, but that go unused because of a difficult user interface. Networking would be used to provide additional capabilities, particularly integration among disparate varieties of office equipment.

For example, today one might print a document, run it through a copy machine, manually distribute some copies, fax others, and send yet another by courier. Instead, under the new, intelligent scheme, one might "print" the document directly to a copy machine, complete with instructions on who should receive copies, and to a fax machine (or server) for transmission when rates are favorable.

By my count, this initiative is actually the third wave in Microsoft's increasingly grandiose vision for the computing future. Wave 1 came in Fall 1990, when Bill Gates announced "Information At Your Fingertips," his vision of a multimedia future in which everything one could possibly want to know about anything represent-able in digital form would be readily accessible.

Wave 2 followed about a year and a half later, with the "Windows Everywhere" strategy, which was designed to put Windows technology on a range of computing devices ranging from small hand-held units to desktop PCs, to RISC-based workstations, to enormous multiprocessor desktop servers.

Wave 3, Microsoft at Work (MAW), unites Waves 1 and 2 in a less visionary but much more pragmatic approach.

MAW architecture The MAW architecture consists of five major components: A real-time, pre-emptive multitasking operating system; messaging and interactive communications; rendering technology that promises to achieve visual consistency across display screens, printers, copiers, and faxes; the Windows GUI; and software that will allow the desktop PC to function as a hub for information flow, control, and distribution.

Let's examine each component in more detail.

Operating System This is where Microsoft's biggest technical challenge lies. The MAW operating system must provide true pre-emptive multitasking, and must be economical in its use of RAM. On both counts, the Windows 3.1 that you and I know and love (and sometimes hate) strikes out. However, the company has in the past delivered a "small footprint" version (Tandy's Visual Information System, discussed here in the December 1992 column), and has undoubtedly learned a few things since then.

Other goals for the MAW operating system include modularity and extensibility, basic support for existing Windows API (Application Program Interface) calls (to minimize training that would be required by developers), and a PC-hosted development environment.

Communications The MAW architecture will support two basic forms of communication: message-based and real-time. The message-based component will be used for enhanced Email type functions, such as a PC-based message-management system that would allow you to sort through Email, voice-mail, and faxes via a single inbox.

Regardless of format (text, voice, fax), each item in the inbox would be identifiable by sender and other information, thus allowing easy prioritization. Other capabilities include "read-only" documents that can be printed, but not edited, as well as fully editable documents. For example, a "fax" might contain both a bit-mapped representation of a document and the complete set of text and graphics objects of which it is composed. The system will also support data encryption, compatibility with existing fax machines, integration with MAPI, Microsoft's Messaging API, currently used primarily for Email.



Technology Area Company Systems, system software, applications, peripherals Microsoft, HP, Intel, Toshiba, WordPerfect Corp.

PC Based Telephone Management Active Voice, Centigram, Dialogic, Octel LCDs and Touch Screens ALPS Electric Telephone and WAN Communications AT &T Easy Link, Bell Atlantic, BT North America, McCaw Cellular, MCI, Motorola, Sprint, US West PC-based fax Modems, Cards, and Software Brooktrout Technology, Cardinal Technologies, The Complete PC, DataRace, Digital Communications Associates, Delrina, Digicom, Phoenix/Eclipse, Expervision, GammaLink, GVC Technologies, Hayes, National Semiconductor, Nuko, OAZ Communications, Optus, Pacific Image, Practical Peripherals, Smith Micro Software, Supra Corp., Thought Communications, U.S. Robotics, Inc., Yamaha Corp. of America, ZOOM Telephonics PC-based OCR Software Caere Corp., Calera Standalone copy and fax machines Canon, Minolta, Mita, Murata/ Muratec, NEC, OKI, Ricoh, Tokyo Electric, XEROX Network-based fax servers Castelle, Cheyenne Software, SofNet, VMX Hand-held computers /communicators Compaq Computer Corp., Casio Telephone systems (PBXs, etc.) Ericsson, NEC, Northern Telecom, Philips, Rolm Ingetrated circuits and chip sets Exar, Casio, National Semiconductor, Rockwell Intl., Sierra Semiconductor, Toshiba Test products Genoa Technology


The other form of MAW communications is bidirectional and interactive. For example, as in Microsoft's current Windows Printing System, a printer could provide aural and on-screen feedback about machine status (paper jam, paper out, toner low, estimated time of completion of current job, etc.). Rendering The rendering component of MAW would provide a consistent imaging model on all computing devices. This amounts to using Windows' Graphics Device Interface (GDI) for imaging, and TrueType fonts for text. Presently, GDI calls must at some point be translated to a specific device format, be it PostScript, PCL, Group Ill fax, or what have you. A single imaging /font model would be desirable on numerous counts; however, the present incarnation of GDI-in Win31 (Windows version 3.1), any-way-is insufficiently rich to accomplish everything that other, more robust solutions (particularly PostScript) can do. The enhanced GDI in Windows NT does provide PostScript-like functionality, but NT has even higher resource (CPU/ RAM) requirements than Win31. It's hard to imagine a standalone fax machine running NT. In addition, the installed base of PostScript devices and applications are not things that Microsoft will be able to overcome overnight, if ever.

Device drivers-and their attendant development and maintenance responsibilities between GDI sources and output devices are likely to be with us for a long time.

GUI Microsoft's intent in this area appears not so much to be intent on making the Win31 GUI canonical as in making GUIs in general ubiquitous. For example, people would probably make much greater use of advanced features of today's standard office telephones if those features had some sort of visual representation and prompting. Look for the appearance of large LCD-based touch screens, whose contents vary extensively according to task, on all sorts of common office equipment.

Desktop Software All the components of this vision will come together at the individual user's desktop. By means of the technologies described here, the individual user will be able to control voice mail, Email, and faxes from his or her desktop; system administrators will be able to manage phone logging, software updates, and maintenance requirements from a central location; users will be able to keep portable organizer-style PCs synchronized with desktop and network servers; and users will ultimately gain more control over how they communicate, with whom, and when.

Vision Microsoft has developed a compelling vision of the office of the near future. It is also a comprehensive vision, much more so than anything promoted in recent years by Digital, HP, IBM, or Sun. As reported here last time, these big companies have rallied around various efforts at cloning or circumnavigating the Windows API. Meanwhile, Gates and Company have expanded the scope of that API by developing agreements with many major providers of telecommunications and office equipment. If I were a strategic marketer for one of the big four, I'd be worried.

But as a consultant and user, I'm excited. I'd like to be able to use my PC to control my copy machine, my fax machine, and my telephone in an integrated manner. If Microsoft Windows is the underlying technology that allows me to do so, so be it.

Realistically, MAW is a vision kind of thing. There are no products that support it yet, and it's within the realm of possibility that there never will be. But I don't think so. I think Microsoft really has uncovered a glaring need, one that was staring us all in the face, if not knocking us upside the head with increasing urgency the past several years.

Several fax manufacturers have stated that MAW-compliant products will be released by the end of 1993. However, it will probably take a minimum of 12-18 months before it will be possible to build up a robust, integrated suite of MAW-compliant devices and applications.

When that happens, our social organism, fueled as it is by information in multiple media formats, will become infinitely smarter and more competitive internationally. When that happens, there will be a fundamental change in the way many of us do business. Those who don't change the way they do things won't be conducting anything-except maybe a funeral dirge.

Digital paper

Less ambitious, but more focused, are competing products recently introduced by Adobe Systems and No Hands Software. Both software packages are in a sense attacking a subset of the problem that MAW purports to solve: efficient document distribution.

Common Ground (from No Hands) and Acrobat (from Adobe) both function as printer drivers and allow highly accurate, electronic versions of documents to be created complete with fonts and graphics.

The two products differ in their underlying file formats, font treatment, costs and licensing arrangements, document fidelity, RAM and disk requirements, and overall product versatility.

No Hands Software distinguishes among three types of font treatments: font replication, font substitution, and font embedding.

Common Ground uses font replication, which works by rasterizing font information at several resolutions, currently 72, 100, 200, and 300 Dots Per Inch (DPI). It stores this information in a file, along with the text and graphics. As long as you view or print at one of the predefined resolutions, you will see a very accurate representation of your document.

Acrobat, on the other hand, uses font substitution, which is based on Adobe's Multiple Master typefaces, which can emulate the height and width characteristics of many Adobe Type 1 PostScript fonts.

(Common Ground supports both Type 1 and TrueType fonts.) The idea is that if a given font does not exist on a recipient's system, Acrobat will maintain line and page breaks by creating, on the fly, a multiple master typeface that mimics the original typeface.

CG's font replication provides more tation; Acrobat allows smaller file sizes. (Microsoft has defined a system of font embedding, which allows end users to embed fonts in documents in one of two modes: read-only and read-write. A read-only embedded font may be transferred as part of another document to another machine and used in that document only; a read-write font may be transferred in one document and installed for use in other documents. However, font vendors have expressed resistance to using this scheme, which could easily promote an already high incidence of font piracy. Currently, the only product I know of that supports font embedding is PowerPoint 3.0.) To view an electronic document created by either Acrobat or Common Ground, a special viewer is needed. Adobe is selling its viewer in single-user quantities for $50; No Hands is giving a mini viewer away for free (via BBSs and on-line services). The company is also selling a more capable viewer for $189 list.

Both products function essentially as the final stage of a unidirectional publication process. It is necessary to run a "source file" through a "document compiler" to produce a distribution file. In addition, Acrobat files can be post-edited to provide hypertext links and other features. The problem is that the one-way process ensures inefficient document maintenance. In other words, if you add hyperlinks to an Acrobat electronic document and subsequently need to update the source document, you would need to re-run the compiler and add the links again.

Acrobat functions as a limited subset of the PostScript programming language, in which strings are embedded in lines of program code.

This makes text searching difficult and inefficient. (In fact, the first version of Acrobat has no search function at all.) Common Ground stores text in a more compressed format for efficient searching.

CG is currently available for the Mac, with a Windows version scheduled for release about the time you read this. Acrobat has released both Mac and Windows versions, and plans to release DOS versions.

Because of the one-way authoring process, I find both products severely limited. In addition, Acrobat's pricing structure is less hospitable than Common Ground's. I would consider using Common Ground for small, limited-distribution projects--e.g., a README file in a software release-but neither is really suitable for large-scale information-development and distribution projects. The concept behind these products is valid; what's needed is a more universal solution. It "HUH ?!"




Top of Page

PREV.   NEXT   More E-N articles HOME