The mission of the District of Columbia's Office of Information and Telecommunications Systems (OITS) should be nicely expressed in only three words: "Make It Work".

Why? "Because we care".

• Buy one new computer (Pentium 200 class).
• Optionally, buy a phone line - a T3.
• Optionally, buy one Ascend MAX TNT multiprotocol WAN access switch.

• Install Linux on the new computer. It becomes a Linux Internet Host, capable of handling up to 64 concurrent users, each of whom can execute tasks simultaneously, while logged in, or as tasks called automatically. This includes FTP and Web servers. Linux is free - no cost whatsoever, other than roughly $40.00 US for two CD-ROM disks and the excellent and very-thick users's-manual.
• Add users to the Linux Host.
• Attach the T3 line to the Ascend MAX TNT, and connect that to the Linux Host.
• Aim local existing computers at the Linux Host. Computers which are or can be made TCP/IP-capable (internet 'ppp') will be aimed at the Linux host, by dialout modem if necessary, but preferably through LAN.
• Dial the Linux Host into the Internet.

At this point, this office now has a computer that is connected to the Internet by a very fast high-capacity line. All authorized users can access this computer via dial-in modem or through the Internet via 'telnet', 'rlogin', or 'ssh' encrypted connection. Many authorized users can access this system via their connections through the LAN. This will back up the existing WAN. The Linux Host can serve webpages, perform mass-mailings to citizens with Internet access, distribute files, permit document creation or revision, etc. Note that even antiquated 286-processor machines can access the Linux Host through ordinary dial-out modem in the same way that they could call BBSes. No new software should be needed.

Repeat as necessary.

• Phase out old computers, and acquire new computers, under the following strategem:
  • Assess and anticipate needs -
    • Users who need only e-mail and minimal word-processing capability can keep 286-processor computers in the interim.

    • Users who need one of the Microsoft Windows operating systems for their work, portions of which may be mandated by outside agencies which require output in proprietary Microsoft formats, or which may be done with software which is provided by outside agencies and is available only for the Microsoft operating systems, should be provided with systems of at least Pentium-100Mhz class. There may be no particular need for extremely powerful computers and it may well be that computers in the high-speed-486 classes (such as the AMD5x86 133Mhz) will suffice. These computers are comparatively inexpensive and are well-tested and data on expected mean-time-to-failure is readily available, permitting accurate projections of replacement schedules.

    • Users whose work does not specifically require the use of Microsoft software should be provided with 486-class (or better) computers with the cost-free Linux software. Linux is a very standard Unix-type operating system, and possibly the vast majority of professional programmers could easily write special applications to fulfill any reasonable need. It is quite likely that various outside agencies and vendors providing custom software could provide it as Linux software. Linux also comes with many features which are available for other operating systems only at considerable cost.

    • Users requiring top-end hardware should be provided with Linux. Linux/Unix has extreme support in the engineering and scientific communities. There are vast InterNetworked software and distributed-processing resources available for Unix/Linux at no cost or at extremely minimal cost. NASA alone offers a vast array of software under technology-transfer protocols, and they may waive most fees for transfer to local government such as the District government, and this does not even begin to address the bewildering array of free software available from the military through technology-transfer and spin-off authorities. The multi-user, multitasking and distributed-processing capablities and in particular the inherent client-server capabilities of Linux/Unix cannot be overstressed. For instance, a secretary's machine might be utilized for concurrent web-serving, mailing-list processing, database operations, and batch-print queuing, and the secretary composing a document would hardly notice any performance hit. Linux also can store and serve MS-DOS/Windows files, and with SAMBA can also deal with Windows NT.

    • All computers should be, where possible, purchased from local vendors who will custom-build to specification from standardized off-the-shelf components, subject to the proviso that warrantees can be secured, generally a standard contract with the origian-equipment-manufacturer. Extreme cost-savings can result from appropriate system configuration. The average secretary has no need of a 300MHz Pentium MMX system, and an AMD or Cyrix 133Mhz processor would probably suffice, and is available for approximately one-tenth of the cost of cutting-edge technology. However, common sense and the antipated need to upgrade should require that whichever class of processor is used, the motherboard upon which it runs should be state-of-the-art. Thus, a computer retired from one service could be upgraded to the highest level simply by swapping the processor for a better one, instead of needing to upgrade the entire system.

  • Mobile Computing/Communications Prioritization
    • Mobile Computing/Communication resources must be anticipated. Increasingly we see a combination of market and technological developments which have the effect of increasing the amount, quality and pervasiveness of mobile digital communications systems while decreasing the cost. The same can be said of mobile computing, particularly as high-bandwidth digital mini- and micro-cellular non-wireline systems are emplaced nationwide. At the time of this writing, for roughly $1500.00 one can have a handheld system that not only has a fair amount of raw processing power, but conforms closely to the Networked Computer mode, outside of the prerequisite of running JAVA. The Apple Newton Messagepad 2100 in this example is combined with a wireless modem and it can then send and receive e-mail or for that matter surf the Web (including encrypted-secure IntraNet) via a built-in web-browser capable of a 16-level greyscale display. Great efficiencies could be realized through centralized distributions of visual or textual information, delivered on request to the civil employee who had need of them. Imagine a cop or for that matter a plumber, who can see the "lay of the land" before he gets where he's going - and who can, once he gets there, have every available record delivered to his eye in any format, from any angle. More advanced (and more expensive, but not prohibitively so) technologies are available including "wearable computers" (or "beltcom"), which are voice-activated and have hands-free head-mounted displays. As technology advances these will become increasingly common, and much less expensive.

      Present District OITS policy is scarecely respondent to the technology of the early 1990s, and any simple attempt to merely bring the District's comm/comp technical base up-to-date is doomed to the greek-classic failure of Sisyphus, forever condemned to push an immovable boulder up a cliff of soft mud. It's essential to be forward-looking particularly as regards office automation; failure to proact with a sufficiently grand vision in tight focus will only leave the OITS struggling in reactive crisis mode.

    • Multiple Distributed Networks philosophies cannot possibly be over-emphasised. A combination of dial-in wireline, dial-in non-wireline, wireline LAN/WAN, non-wireline LAN/WAN must be considered, and where possible, future developments "Ubiquitous Computing" should be anticipated in all developmental strategies.

This page and any visions, strategies, or mission-concept formulations expressed herein are the intellectual property of TJH Internet Sole Propriety and are copyright (copr) all rights reserved as of 30 October 1997.