This chapter should discuss remaining topics regarding the system in general and also other related issues.
 

7.1. System further developments

What required system enhancements do we envision, what should be done next? Well, though question. Of course we can add a bunch of new features like a direct inter-entity stream facility for example or maybe an improvement in executable server side code, a new communication handler that improves message routing and specific facilities etc.

The main purpose of the system is to be fast, robust and simple to use in the same time providing a must-be bunch of needed facilities and the provision of openness allowing for new ones to be easily integrated. In this sense, the main improvements should regard the area where the system is used in particular, in a certain application that needs improved performance in certain system areas.

For example Talking Objects makes full use of group and broadcasting facilities. This makes it imperative for a large topology solution to implement a communication handler that optimizes those facilities, by using multicast IP for example.

The above example is relevant for the system design issues. Optimizing issues should arise only in the case of real usage it. For any comments or requests please contact the author. We'll be very glad to support any of your submitted improvement or new facility.
 
 

7.2. The net of the 2000s

The next millenium's computerized networked world. It sounds scary but what do we actually expect from the near and long term future computer networks? Our vision is that of a networked system composed of virtual entities, objects, interconnected by means of virtual communication paths, inside a global medium-alike environment that offers several basic standard facilities including at least at a base level also those that we name today directory and lookup services, authentication, communication security (maybe encryption) etc. Each device in the real world is to be integrated somehow into this network by having a unified interface for a generic virtual device entity manipulation. Each running task or application or even software potential and computing power should be also integrated. The network is seen as a living being, running at its maximum capacity and providing a global integrated information and control highway. In this frame we mean by device also any actual physical device, like a cellular phone for example or our home washing machine. There are times ahead when some hacker will try to overcook our eggs by changing the microwave parameters inside our home oven.

Being scary enough, the above vision is becoming part of our life day by day as time passes by. One of the major points gained towards it is the market transition from the highly sophisticated technical buyer towards the low-end mass consumer. Once in each 2 years the hardware in personal computing halves its prices. If you buy today any piece of PC hardware you'll be able to buy it at half a price in two years. This is one of the only domains where you just have to wait and do nothing for things to become amazing cheap. Another major point is the impact of the simple and initially naive World Wide Web technologies into every human's life. Although this is not, at least the current look of it, the technology of the future, it might grow towards it, becoming maybe the ancestor of it. New approaches like VRML, Java, Virtual Communities etc. bring us every day one step forward. Will this make humans better?
 
 

7.3. Small refreshing monologue on system scaling

This section is not written with an academic purpose. If you're looking for such a type of refreshment please stop reading it here. It was written just as result of an email discussion had with a very nice well-known net and OS expert, I had some time ago.

How do virtual distributed systems scale? They don't. For answers maybe we should look at some good Systems Theory books.

In the second world war, the Germans weren't able to build the radar in the first place because they always tried to simply scale a prototype. They found out that some systems don't scale.

This should be a lesson. A system designed to be a prototype maybe suffers from the same kind of 'disease' as predicted by Heisenberg's rules of not being able to determine at the same time the speed and the impulse of a moving particle with enough good precision. A prototype is a prototype. In some cases scaling works. Strange enough in most cases it doesn't although it seems to us that it does.

System design should always start by having in mind constantly a working ideal that keeps the spirit alive ;) and each move should be made, even in the localized, limited issues that arise during the design and implementation phase, towards reaching this ideal. If the ideal is building up a global virtual system composed out of many entities interrelated (I won't use 'distributed system' anymore) than every single piece of design should be done towards this goal.

Another issue would be NOT to mix the two phases of design and implementation. Once design issues are set they should not be changed and the best effort should be done in order to construct a working system that complies mostly with the specs. What I mean is NOT to let the current hardware limitations for example establish the real important design issues.

And as always a good example won't be of any harm. The perfect ideal virtual system is the real world as it is. Some might wonder and stop reading here. But no. The system that we call 'real' world is a full ideal virtual system composed of evolving entities interrelated. But how is this world created? We don't know. But we can look at least at the effects if not at the core mechanisms and try to derive our own mechanisms that maybe comply in producing same effects, that is allowing a unlimited number of entities composing the medium and also unlimited type of interrelationships.

As said we cannot penetrate the core of the system but we can look at some characteristics. Communication in whatever form it might be represents one of the main ones. Communication brings about rightaway the next issue, identification. Because of the nature of this medium that we live in there is another very important characteristic that appears also in any other system but maybe under some other name, localization. Localization determines communication and is based on identification.

The above has also a deeper mystical meaning. The world is created using the Word, which is a symbol of identification. The alphabet in itself was used to implement the Word, which brings us to communication. Through successive mystical contractions it finally descended (localization) and got to be...

See the file INCONVENIENCES-BUGS