A Mutable Log

A blog by Devendra Tewari

Project maintained by tewarid Hosted on GitHub Pages — Theme by mattgraham

Perspective on software

We have just rung in 2003—2002 is already a good riddance to some and a nostalgic year to others. Several things have happened in the last few years in the software market. We have seen tremendous success—as in the case of the Internet—and tremendous disappointment—as in the case of stock value depreciation of several Internet age software companies.


The software world is in flux—again. A decade ago, only the bigger companies could afford to use IT. Today, every company and household does so. There is only one problem—software is still very expensive. Every two or three years, most companies and households have seen the need to upgrade most software—from the accounting package to the operating system (OS). This has created the impression that software and hardware upgrades are a necessity.

That is slowly changing and giving way to a new awareness. Now that basic IT infrastructure needs have been met, most companies see no reason to be on a constant upgrade path without a decent explanation of the good new software will do to the bottom-line, and bottom-lines are hurt by a bad economic climate.

Market Forces

Let us take a look at the software technology forces in the market vying for success—touting innovations in reliability, security, and use.

software technology forces in the market vying for success

Legacy systems made in COBOL and running on older mainframes still run in several big name companies, but these have mostly been replaced by Unix servers in most companies throughout the past two decades. That brings us to Unix servers—most of these run some variant of Unix and use specialized hardware to do anything from data crunching to print serving. Most of the Internet infrastructure runs on these sturdy and powerful servers. Linux—a Unix variant—is a recent entry in the systems software scenario but has reached maturity relatively fast due to its strong roots in the Unix community.

DOS—the predecessor of Windows—started out as a personal OS. It wasn’t meant for server computers. DOS machines can be credited to have hooked a whole generation of users to the personal computer. The first commodity OS was born. That laid a splendid foundation for Windows which continued the process of enticing virtually every human being on the planet to the personal computer. The visual metaphor provided by Windows is now commonly used in almost every application—whether it runs on Windows or not. Linux has been working to provide a cheaper alternative to Windows.

The last decade saw the emergence of Java that brought a paradigm shift in how software is developed and consumed. Software can now be consumed over the Internet. Java has since evolved. It is now used on servers as well. Its main advantage seems to be a virtual machine specification which, if implemented for a particular platform, makes consumption of any Java software possible. Java gives us the true possibility of making application software a commodity. That brings us to .NET—which provides essentially the same possibility, but Microsoft—so far— has seen no need to provide .NET on anything but Windows.

Software as a Commodity

Software and hardware are turning into commodities—as several industry observers have mentioned repeatedly. This usually is a reasonable indicator of a mature industry. System software by its nature is costly to build but very cheap to spread. The question is—if hardware feature innovation stagnates will we ever need to upgrade system software? The answer to that may be the reason why Microsoft is being criticized for leading every company and home user to upgrade Windows. Microsoft, I am certain, is aware of that and is looking for more innovative ways to make money. Linux, in the meantime, is pushing into the systems software commodity market with a price tag that is hard to go unnoticed for long.


Personal computing is also progressing into increasingly smaller devices. Embedded systems—in which a personal computing device has been married to a modern telecommunications network—ought to ring in a new era of computing, forcing businesses to adapt their server side resources to service this new market. Once again, a personal computing revolution will be driving changes on the business side and not vice-versa.

In the meantime, let us ring in the new year—happy 2003!