Now here's the good news about OS/2

From Wiki
Jump to: navigation, search

by Martin Brampton

The IT press stands accused of downplaying the importance of OS/2. Martin Brampton presents the case for the prosecution.

Like Mark Twain, users of OS/2 have been perplexed by press reports of their death. OS/2 has existed as a live operating system for a decade now, providing a robust way to release the Intel architecture from the historic chains of DOS.

Reliable figures are hard to come by, but for a system that is often pronounced dead OS/2 seems to be doing surprisingly well. It is claimed that more OS/2 licences than MacOS licenses have been sold to date. In 1966, incredibly in relation to press coverage, OS/2 sold more copies than Windows NT. There are many organisations that continue to depend on OS/2 for vital business functions and continue to purchase new copies to extend their operations.

Of course OS/2 hasn't lived up to the highest hopes of its progenitors, IBM and Microsoft. But such is the hype surrounding computer developments that many products have a long and useful life without ever achieving what was hoped for them. Unix has never achieved the universality that some anticipated, yet it remains a trusted system that provides sterling service to many organisations.

Making steady progress through new versions that add functions in small increments and preserve the viability of earlier applications is perhaps not the way to maximise news coverage. Nor is OS/2 helped by the apocalyptic style of press coverage which tends to see everything in terms of a battle of giants for total supremacy.

Yet for those of us seeking to implement solutions to everyday problems, life is more prosaic. We're simply trying to provide real value over a period of years, using expensive computer technology that is apt to lose its value faster than we can install it. In fact, far from being our wish, it is our fear that key suppliers will become dominant, a situation which often leads to complacency and unresponsiveness to customer needs.

There are many ways to achieve our goals, and it is not my aim to criticise the alternatives. Just to report that a sizeable group of people have found OS/2 an effective solution and a platform on which to build. For the decade of its existence, OS/2 has preserved the viability of the earliest applications. For more than half of its existence, it has been a 32 bit system which fully exploits the Intel 386 (and beyond) architecture - the first Intel architecture to be designed knowing that its main use would be for personal computers.

And many developments need at least those timescales to be effective. In the flood of innovation, it seems to be forgotten that software often has a surprisingly long life. It is the root of the millennium crisis that there is still value in software that may have been in use for twenty years. In the everyday world, there is a mix of software, some of it freshly born and some of it built over very long periods.

The commercial pressures in which I work are focused around continuity of service and ease of use. For example, we aim for 99.9% network availability. We chose OS/2 to permit a long enough period of use to justify devoting effort to refining our ways of operating. The base system, combined with IBM's Restricted Workplace Shell has enabled us to build workstations that provide a graphical environment which derives totally from the network and the identity of the user. Each workstation is the same from a software point of view, greatly simplifying administration and enabling roving usage even across different locations. Working that way has been helping us to provide reliable services at the same time as controlling support costs for the past several years.

The future increasing looks to be moving towards thin clients of some kind. OS/2 is already well positioned in that respect, as the first operating system to integrate Java support. It is winning praise from Java developers as a robust and efficient platform. OS/2 will operate as a client for Citrix's technology to provide applications from a multi-user Windows NT server - a method of running applications that is rapidly winning converts.

Further out, the "Bluebird" project is generating considerable interest as a version of OS/2 that can load from the network and provide a robust, flexible environment to run applications on an Intel platform. Bluebird targets a spectrum of uses from access to mainframe applications to running the emerging Java software.

Oh, and let's not forget servers. There is real competition in the server arena, especially between Netware, NT, OS/2 and Unix. Perhaps this shouldn't be seen as a single market, as it is increasingly splitting into different kinds of servers with different servers matching more less well. In any event, OS/2 has a share of this market, and research companies are now indicating that sales will grow over the next few years.

So can the many people using OS/2 hope for the occasional press report that talks about OS/2 developments without telling us our system is dead?