Why Write Free Software?

The GNU project arose from the Free Software Foundation, the brainchild of Richard Stallman. I urge you to follow these  links, as there is much documentation on the motivations for the  generation of free software, as well as copious examples of it in action.

Freeware is typically written by enthusiasts around  the world. These enthusiasts will often work alone, or as loosely co-operating groups, across the Internet (they may never actually  meet). In fact without the Internet much of the freeware in existence would never have been written.

Some are students, many work in commercial enterprises,  and some are even sponsored to participate by the organisations they work. The motivation is hard to pin down, peer acknowledgement  must be one, the pride of recognition - some of you will recognise  the name Richard Harris, the author of Pegasus, many more will  be aware of Linus Torvalds.

Some contributors are just benevolent, others enjoy  writing the tools they create and wish to share. The key factor we must understand though, is that it clearly does happen - it happens a lot, and the output is not only sustained, but of high  quality, timeliness, and often appears to be a more effective software generation methodology than traditional commercial techniques.

 One perspective on this is that there are people  who consider software development to be just like any other science. Research and development is built upon earlier discoveries, as a skyscraper is built upon a known foundation. If software innovation  were totally proprietary, then everybody would design their software  from base principles. This is not how you put a man on the moon.  If someone had managed to copyright Hydrochloric Acid there would not be much of a plastics industry.

If security is a concern, then you can rest as easy  as many other organisations. Many corporate firewalls are actually Linux on the inside, hidden by a web interface, and manufacturers  of some palmtop computers are suggesting that Linux could be the basis for many future handheld devices. This is in no small way  due to the fact that a palmtop's operating system is one of the most significant cost issues in their production.

Ironically, the collaborative development community  across the Internet, which spawned freeware, appears to be producing  more stable code, more quickly, than any commercial enterprise before it. At the same time it's providing a global skill base which is not tied to any single organisation.

 It is the reliability and stability of freeware that  is driving organisations to use it in e-commerce applications. With the advent of the latest mainframe and highly scalable versions  of Linux, this reach is likely to spread into more traditional commercial areas.

In the last year there have been some very notable high profile Linux installations on mainframes for eCommerce deployment,  and several highly parallel clusters created for supercomputing applications. In fact I understand IBM have just committed to a  mainframe running Linux for Warwick University (my old stomping  ground) for research into the next generation technology of the Internet.

How is it Supported?

 Because freeware is vendor neutral, anyone can  support it. There is no supplier that you have to contact for  a fix - bugs are found through peer pressure and posted on the  Internet with incredible speed. And there is always the option  for customers to fix a fault themselves if the situation is serious  enough. Try that with commercial code! There is reassurance in  knowing that if it came to it, you could resolve your software  problems without relying on others!

There is also tremendous power in the ability to self-customise your freeware environment to your own exact requirements if need be.

 Of course not all companies will have the technical skills and resource in house, and this is being addressed by organisations that can see the opportunities. First on the scene in this area  were small consultancies and the Linux distribution collectives such as Red Hat, Debian, Suse, etc., but now major vendors are  pouring huge sums of money into this area including IBM and HP.

Why is IBM getting involved?

 As margins on commodity IT hardware dwindle, IBM, HP, Sun, Oracle and others are increasingly mutating into service-focussed companies.

Software is probably heading the same way too! Most  software vendors are examining charging schemes that rent out  software as a service and include concepts such as ASP or charge  by usage. In fact, there may well be as much a revenue opportunity  in freeware as in commercial software. As these new charging schemes  take bite it is going to be ever more apparent to companies what the ongoing costs of their IT operations are - and what the alternatives are.

What is the Future?

I don't pretend to be much of a fortuneteller, but I understand enough about business to know that in order to thrive  you must grow revenue and/or reduce costs. Software costs and the computing life cycle costs in terms of purchase, support and  upgrade are increasing, at the same time as viable freeware alternatives  are becoming available.

It is true that some applications will never be  free, but at the horizontal and appliance level, as freeware becomes more commercially acceptable, its take-up is inevitable. As an industry we simply have to work out how to manage this opportunity to our advantage.

Barry Fairburn