by Set Lonnert
Java will soon conquer the world -- some say. Microsoft, the real dragon, will soon swallow Netscape and anything in between -- others say. Well, then either path could be entered, or then none of them.
In my opinion, Netscape will have to revise their strategy drastically if they want to stay alive. No huge programs or plug-ins. They will have to do a lot better than that. Site-constructors of today have to make some rather unnecessary choices. Either stay with rigid Netscapeism, or make browser-sensitive pages, i.e. dynamic pages that can utilize several browsers. In either case, the visitor is obviously a big concern. Most visitors make their visit by Netscape-browsing homes, and consequently Netscapeism has become the natural outcome. But then you also may have the broader perspective and seek a little more general approach like -- any browser will have to do.
Microsoft, close behind Netscape, is making their own browser available to the community of PC-owners (running Windows of course) producing another soft-soft technology we do not want.
Netscape chose the wrong strategy thinking that huge programs were still a good idea. They are not, not anymore. Apple long since, and then Microsoft, have engaged in ways to embed object oriented programming (and object oriented thinking) into their operating systems. Why did Netscape fail to see this prior to getting involved in enhancing their browser (Spry seems to take this ad notam)? Netscape could have gained time instead of money. What will they do in the long run? On the other hand, perhaps they making one, at this very moment. Then, naturally, there is an issue of the plug-ins. Instead of just making objects plugged in we ended up with a monstrous product created by huge programs. Java is today paradoxically a plug-in in Netscape. It will soon be reversed.
These are some of the problems we are facing today. What will the future bring? There are at least two things as far as the most promising replacement technology, Java, goes; Network Computers (NC's) and JavaOS.
Network Computers are coming to life with some JavaOS installed. They will cost about $500, or so they say, but in fact it could very well be the double. The NC's have a very different strategy to networks than personal computers, browsers and built-in protocols. It will not just be another browser on a new machine. Not by far. The game-industry and the telephone-industry will first be involved in these developments. Then other industries may very well follow and the much talked about idea of connecting electronic chips world wide will probably be a fact. You can expect your wristwatch to run JavaOS (embedded API) as well as your expensive JavaStation (with not yet written API for hardware drivers).
Sun Microsystems demonstrated a machine with the (still rudimentary) AWT run directly on hardware with its own implementation. It runs twice as fast as AWT hosted on the same hardware under Solaris. All in all, it can run with as little as 4Mbyte of RAM and 4Mbyte of ROM. Maybe NC's will have a big cache, so it also can store some software when not connected (downloading the JavaOS every time you start would be a little bit cumbersome). If they run VRML they must have some (RAM) store for direct access. It has no delays due to a JIT, but instead the byte codes may be directly executed. But then new chips are coming (JavaOS is expected to run on at least ARM, CompactRISC, Intel X86, NS486, PowerPC, microJAVA, microSPARC, picoJAVA, SPARClite, all according to Sun).
Well, what will happened to the old browser then? Will it die? Yes, I think so. The ability to change protocols (ftp, http, or whatever), lies the foundation for whatever comes after the Web or at least rise above e.g. unsecured manipulations. The Netscape Navigator, as it is today, doesn't have the capability to change much. Java does, and hence whatever is built on Java, such as a future browser (i.e. HotJava).
[Upp] [Börsen] [Referesnböcker]