By Al Williams
When a person needs to make a decision about what software product to use, why is it important to have a variety of options? This is a situation where diversity can be a strength. It is good to have the freedom to choose from multiple solutions, realizing that freedom to choose does not necessarily imply that products are free. A decision maker needs to focus on the business need at hand and make technology decisions that support business strategies. We need to be able to control our destiny, and not let vendors unduly influence our decisions. Viable options better position us in a bargaining environment. We need to maintain the ability to change if or when needed. And we must maintain budget independence (no one vendor can control).
What choices do people have when they need to make a software selection decision? Let's look at that question and break it down into specific cases: servers; desktop OS; office productivity suites; and browsers. Looking at the situation that way may give us a better understanding of where we have reasonable choices.
There are quite a few server choices available: Microsoft Server 2003; Novell NetWare; OS2; Linux; Apple OS X Server; IBM AIX; Sun Solaris; and IBM z/OS. Of these Linux is used quite a bit here at Penn State. Linux is the back end server for:
Red Hat and Suse are two companies that provide Linux products targeted for the enterprise marketplace. They sell the Linux system as well as support and automated bug fixes. AIX is actually gaining share in the Unix server market, and z/OS is prevalent in the enterprise commercial market. The latter is being helped with the availability of Linux on IBM mainframes. Apple OS X server is a good product to watch. It has recently gained market share in Europe. A subset of the server systems is the web server. Here you might choose from the following: Microsoft Internet Integrated Server 6 (IIS 6); Apache; Sun One; Lotus Domino; and Netscape. The most prevalent web server on the Internet is Apache with about 60% market share.
In the desktop operating system market there are several choices: Windows XP; Apple OS X; Linux; Proprietary UNIX (Solaris, AIX, HP UX...); and free Unix (FreeBSD, OpenBSD, Darwin...). Apple OS X is a product that I believe is really worth considering. Linux bears watching and is becoming a contender. Several vendors have launched desktop initiatives based on Linux (Sun, HP, Red Hat, and Novell to name a few). These are noteworthy efforts that are gaining traction.
There are a lot of choices for office productivity software: Microsoft Office; StarOffice and OpenOffice; Corel WordPerfect; Lotus Smartsuite; Apple Works; GoBe Productive; and KOffice (to name a few). ITS struck a deal with Sun and distributes StarOffice on the student software CD (http://its.psu.edu/studentorientation/pacits.html). OpenOffice is the open source derivative of StarOffice and is freely downloadable from http://www.openoffice.org/. Most of these products tend to be pretty good at interchanging documents, presentations and spreadsheets.
For web browsers we have some excellent alternatives: Internet Explorer; Mozilla; Netscape; Safari; Opera; Foxfire (from Mozilla). I list Mozilla separately from Netscape as it seems to be developing into an interesting open source desktop platform offering email and calendar bundled in. Mozilla is freely downloadable from http://www.mozilla.org/. For email clients you can choose Microsoft Outlook, Outlook Express, Mozilla, Thunderbird, Ximian Evolution (looks and feels like Outlook), OS X Mail, Eudora, and many more. Mozilla and the Mozilla Thuderbird spin off are worth considering. They work very well across the Linux, Windows, and Mac OS desktop operating systems, and you can even copy settings between the operating systems.
Several of the options mentioned are open source products. This brings up the question of whether we should be favoring open source as a solution. Before we can answer that, we need to understand what is open source? For our purposes we use the following definition:
Open source is an application, program or utility where:
Open source is a popular revolution, and some good work is being done. The idea is to share development resources (a Stone Soup Model).
More eyes, i.e. more people, can help find errors. Problems and security holes are quickly fixed, and improvements can be made freely for the greater good. This does not imply that open source is free. You may need to have programmers to support open source products you adopt. You will need to have an appropriate programming infrastructure to be a participant in open source development. Vendors can resell open source products where they add value. Indeed there are quite a few vendors starting formal open source initiatives:
The March 1, 2004 article "The Myths of Open Source" from CIO Magazine gives some good reasons a business might choose open source products (http://www.cio.com/archive/030104/open.html). The article addresses issues such as the cost of open source, savings potential, and support issues. We actually use several open source solutions within ITS:
We also provide an open source mirror web site (http://carroll.aset.psu.edu).
If open source is a reasonable choice, what are some of the products, and how do you find open source offerings? Those are good questions. The products are very numerous, so I will just give some examples here, and also give some pointers to where to look for open source products. Here are some fairly well known products:
Those are interesting examples, but how do you go about finding what's available? Here are some references that you can browse:
- Apple Darwin
- Open Source Directory
- Unix apps for Mac
- GUI for Fink
http://sourceforge.net/projects/finkcommander (you need to install fink first)
So now we are down to the question of what should we do? Should we:
The answer to those questions is "yes" depending on the situation. Certainly ITS favors Open Standards based solutions. In my opinion it is key that we strive for interoperability. We want to be able to make decisions based on our business needs. Source managed solutions can be good (like our ANGEL project), and open source is a good option where the products match our needs. I believe we could and should work together to influence our vendors to be responsive to Penn State needs. We have done this before in order to address our concerns at the source; try to influence product direction; try to influence interoperability; and to build standards that we can live with.
So we come back to the question of whether there are meaningful choices and reasonable ways to diversify? Yes, I think there are reasonable options and we can work together to realize them by making a conscious effort to look at alternatives. Another option is participation in university-focused initiatives such as:
And a third is participation in user groups that seek to influence vendors such as
By using these strategies we can make software choices that are in the best interest of universities as a whole and Penn State in particular.