A FEW days ago a large US retailer confirmed that it will use a Linux-based solution for its new retail systems in more than 100 stores. Jay Jacobs, a men’s and women’s clothing chain, has officially announced that open source software is an acceptable alternative to Windows NT for running a commercial retail operation. Linux, for those who are not on top of the IT trends, is a free, open standard operating system which happily competes with Windows NT and has actually got a good chance of taking on the Microsoft brigade.
Linux runs on any hardware platform and, except for high-end multiprocessor tasks, has recorded a superior performance over both Unix and NT. It is continuously developed by an army of enthusiastic volunteers who run a special community of users, developers, documentation writers, support people and testers. It is free and support is provided by a number of software companies such as Red Hat. Jay Jacobs’s announcement marks a milestone in the transition of Linux from a hobbyist environment to a fully-fledged, industrially robust software platform. It may have helped that Oracle, Informix and Intel are fully supportive of Linux and have gone on record with their commitment. Many top IT suppliers are tired of Microsoft’s hegemony and will happily support the Linux alternative as long as the new baby does the job.
Open source software has become a way of life for a large number of developers since Netscape released its source code a few months ago. That seismic move rattled the cage of the proprietary-oriented software companies and signalled the beginning of a new era in which the wit and intelligence of the collective programming skills of volunteers will take on the mighty software companies.
The risks are still high. Linux does not work for the large, multiprocessor tasks, and its support base, although enthusiastic, has not quite yet proven to be of industrial strength. So why did Jay Jacobs’s IT people go for Linux?
One of the reasons might have been the number of bad experiences the company had with proprietary software, where the moment you complete the development, you face a constant threat of lack of support. The reasons for ceasing support are countless. There may be a new version of the application coming up, and the supplier will not support the old one. There may be a recession and your vendor has gone out of business. There may have been a shift in the market and the product was discontinued. Finally, it may not be supported on that new machine you have just obtained as part of the carefully thought through upgrade. Many long-suffering owners of e-commerce system in the UK are currently in that boat, where some of the large IT suppliers have changed alliances rather suddenly and the current version of their e-commerce application has suddenly stopped being the centre of company strategy. Clients can be asked to change to a Microsoft solution because Unix is not on the supplier’s list of partners. If the client’s architecture doesn’t allow that, then they are on their own.
Since Linux is simpler to develop for than NT because it has open source code, it has created a hydra of multiple programming heads in various parts of the world. Thus problems are identified quickly, solved and the results shared by posting them in newsgroups. This speed of knowledge growth is staggering and neither Unix nor NT can even start to compete, considering that each of them have a limited number of developers on their teams, while Linux has the headspace of thousands of the cleverest young programmers.
However, this is where the drama starts. Realising that Linux is gathering momentum, Microsoft has just woken up and developed a “kill” strategy. Two memos, leaked on Halloween night from high-level Microsoft cyborgs, have demonstrated that Bill Gates is not going to let Linux steal the show. The memos identify the success of Linux as a function of allowing people to use open protocols such as e-mail or HTML, and therefore forcing Microsoft to live with a commoditised environment. According to the leaked memos, those protocols can be “de-commoditised”. In effect, that will destroy the environment in which many people can develop new applications due to open standard in a widely accessible environment.
Now is the time for swift action to freeze the open standards of key protocols and impose heavy fines on the companies which don’t toe the line. The de-commoditising of open standard protocols would quickly lead to higher prices for networked applications, suppressing competition, reducing choice and, worst of all, having more Microsoft on our desktops, cash tills, TVs, organisers and other future computing devices.
The stellar progress of Linux makes us all question what infrastructure is required to write good code. If a virtual team of worldwide hobbyists can come up with the goods that a bunch of highly paid people in big companies with big company cars couldn’t resolve, then what is the added value of a corporate structure based around programming teams?
Before Linux, it was a commonly shared assumption that anarchy is worse than organisation, and that software written by developers organised in tidy teams and supervised by a manager in a suit is better than software developed by volunteers in a self-organised community with no manager in sight.
If I were an IT manager, I would be very worried indeed. The inflated costs of software we pay today to cover managers and their cars will go out of the window when clients realise that there is an alternative thanks to open source code. Bill Gates should be joining the hackers, or taking a deep breath. If Linux continues its progress, it will not be good news for Microsoft, but it will be great news for the consumer.