Office Wars: Applixware and StarOffice
We humans love glowing boxes. Monitors, TV sets, suns, moons, lanterns, candles—it all goes back thousands of years, when we sat around in tribes staring into the fire. Computers? As long as we're comfortable while staring into the glowing box, it matters little what exactly we're doing. The light is comforting, hypnotizing. Still, some crazy person got this idea that computers ought to have a use, nay, a killer app. Now we're cursed with spreadsheets, word processors, databases, development tools, office suites, mathematical packages... Well, Scott McNealy doesn't want your money (at least not right away), so we might as well make the best of it.
I must admit, the “warring office suites” cover has an air of commercialism to it, but it's not far from the truth. At the same time, it doesn't tell the whole story—Sun and Applix have their minds on bigger things. It is true that on one level, we have a price war in a particular market (cross-platform office suites for Linux/UNIX), but Sun's plans for StarOffice and Applix's ideas for Applixware are fixed on a less immediate horizon. Before we look at the offerings of these respective packages, it's worth the effort to find out where these products are heading.
Sun acquired StarOffice last August and has been giving it away free of charge. It's not open source, not guaranteed to always be free, and it is proprietary software replete with generic licensing nonsense—still, you can now download all 65MB of it without paying a fee. You can also order it on CD for $10 or $40 if you want documentation and support. The question is, where is Sun going with this? A typical cynic might expect Sun to dump StarOffice on the market, drive out the competition, and then start charging for the updates—typical rent-seeking behavior. However, Sun's plan is a bit more visionary. Remember web-based e-mail? Sun wants a web-based office suite.
StarPortal will be the name of this “portal computing” office suite, but since it's first-class vaporware, we don't know how it will be implemented (well, Java). Still, everyone wants to cash in on the falling cost of bandwidth. StarPortal will have many obstacles to overcome (stability, security, accessibility/availability, even the client/server implementation itself), but at least Sun has the size and capital to pull off this kind of project as far as it can go, and Sun owns Java.
Applix, not to be left behind, in fact to be in front with software instead of vaporware, is also expanding into the thin-client office suite market. Anyware Office, Applix's answer to the portable office problem, is an 800K Java applet which allows a user to have office suite access from within any Java environment (Netscape, IE, JavaOS, etc.). You need to have Applixware on your home machine, of course, but if you've got this much, you can set up a home office and contact it from any terminal whenever necessary. It's quite a clever model, one both Applix and Sun are pursuing. One difference in strategy is that Sun hopes to have StarOffice accessible even from PDAs (personal digital assistants with their half-functional web browsers) while Applix, at least for now, is staying solidly on the functional web-browser level.
Another issue is the use of Java. Years ago, when asked for my opinion of Java, I said it was not a serious language, but only a distraction, trendy and not worth the effort. Nowadays, Java has grown in popularity, not entirely on its own merits but because of C's memory management issues and C++'s tendency to produce memory-leaking monstrosities (well, low-level languages do expect you to deal directly with memory). Java has been quite successful on the Internet and intranets and is enthusiastically supported on account of its attempt at platform independence. However, developers must be aware that Java is a proprietary language. The specification is open: anyone can write Java programs, compilers or interpreters, but Sun owns the rights. It's probably not a good idea to become dependent on proprietary languages, and much better languages are available, but Java is nevertheless where our Linux-based office suites are headed.
Now we know where these office suites are going, so how do we choose between them? This assumes, of course, that we are willing to rely on commercial software. It's mildly difficult to get by without using commercial wares from time to time, so if you can, maybe you are a saint, or at least truly clever. Office suites are often the justification for getting a computer, and if you've been in school recently, you know exactly how much of your otherwise pleasant life is spent in front of a word processor. Likewise, science types may often find themselves chained to spreadsheets and need to have a fast, stable package with all the formulas and graphing functions that might come up. In any event, if you want to use a computer with only one OS on it and expect to get something done other than hacking, a word processor is a good idea.