I have spent the last 10 years implementing, using, and advocating Linux for a variety of applications. During that time I have watched the steady progression of Linux, gaining success as a server, desktop, and embedded operating system. The facts are indisputable: Linux is a success and it more than adequately meets the needs of many enterprise class applications and open source operating systems, chalking up wins in both consumer electronics and on the desktop. The next stage of important growth around open source is not in the applications we use but in the complementing applications we use to manage and expand already established software.

The term tool is very broad. It may refer to software used for software development, management, or monitoring. Within these three critical areas I believe that advances in virtualization, systems administration, and rapid application development tools hold the most promise. Virtualization is one of the hottest buzz words in information technology. For some time, storage virtualization has helped reduce the complexity in managing and dividing huge storage devices and storage area networks into discreet useful volumes. The newest trend is the virtualization of the whole machine, allowing virtual servers to span farms of commodity blade hardware or for desktop PCs to run both Linux and Windows simultaneously. Advances in the open source Xen hypervisor (www.cl.cam.ac.uk/Research/SRG/netos/xen/) as well as new offerings from upstart Virtual Iron (www.virtualiron.com) and established virtualization player VMware (www.vmware.com) are redefining the data center by helping to eke out even more value from the low-cost commodity server hardware. One popular theory is that in the data center of the future you will provision software into these virtual server farms rather than run individual networked servers. In hosted services we already see great success from SWSoft's Virtuozzo (www.swsoft.com/en/products/virtuozzo/), which allows for dynamic provisioning of virtual Linux servers. While Virtuozzo's approach is tailored more to toward maintaining policies for the commercial-hosted data center, the key is to have tools that are aimed at maintaining the intricate policies of the complex enterprise data center.

Other than virtualization, systems management is an area that I find fascinating these days. You see, as Linux rapidly came of age, the tools to manage Linux and other open source technologies have yet to match the progress of the platform. These technologies need to be distribution-agnostic and need to reach across operating system boundaries to manage not only Linux, but Unix and Windows.

With that said, I think there is real potential for products like Centrify (www.centrify.com), which helps to integrate non-Microsoft systems with Active Directory because it's helpful in managing and maintaining continuity among a homogenous data center. Another element of management is provisioning; in this space rPath (www.rpath.com) has its eyes on the interesting combination of provisioning and virtualization. Their method of provisioning is to build task-oriented Linux distributions (e.g., Web servers) on the fly and then allow you to either deploy natively or alternatively as an image that can be executed in a session running under Xen. Their technology builds a Linux distribution with all dependencies resolved so that the applications and operating system are able to be quickly deployed. I also like Levanta (www.levanta.com), which provisions Linux systems but is not distribution dependent, giving you the flexibility to choose the distribution you like or to have a combination of distributions if necessary. These types of tools are especially useful because they offer systems administrators a central point to deploy software and operating systems. Once you deploy, your next order of business will likely be to monitor the health of systems both virtual and physical using monitoring technology like OpenNMS (www.opennms.org) and to a lesser extent Nagios (www.nagios.org), which now draws corporate support from IT Groundwork (www.itgroundwork.com).

Finally more development tools that rival Microsoft's Visual Studio and other rapid application development tools must make their way into the hands of open source developers. Earlier this year IBM donated Rational Unified Process (RUP) to the Eclipse Foundation (www.eclipse.org) in hopes of fostering more development for robust Web server applications. For years the developers of KDE have been using the QT cross-platform toolkit from Trolltech (www.trolltech.com) to develop the popular Linux desktop environment. Both of these are good tools, but I recall how, in the early days of the Web, FrontPage became a force to be reckoned with because it put the ability to author Web pages into the hands of most any Microsoft Office user. Now the context and complexity of technologies are quite different but the ideally the impact of the software should be the same. Development tools that quickly expand the potential developers to a comparable degree will help advance open source technologies. This is especially true on the desktop where potential Linux users often suffer from application availability.

In 2006 I am looking forward to the progression of tools that will help to develop, configure, manage, and monitor open source applications. The resulting developments will be the proliferation of the same caliber of tools that Windows and Unix users are privy to. I believe that the open source methodology will lead to a better breed of tools because collaboration between tool makers and the technologies are facilitated by a better level of communication and collaboration then with proprietary technologies.