Современные информационные технологии/ 1.Компьютерная  инженерия

Bulatova M.B.

 

Kostanay State University named after A. Baitursynov, Kazakhstan

 

Conception of intellectual system of utilization computing capacity of Internet.

 

At present there is a tendency of use of Internet generally only as information system. Therefore it is naturally that almost all modern Internet services are aimed on work only with documents and files. E-mail, teleconferences, ftp – are incomplete list of the most popular services, which main destination delivery to the user of information from remote sources.

Now World Wide Web also, mainly, serves for the appeal to information resources, and overwhelming majority from 30000 of well-known Web-servers can’t offer anything except hyper media Home Page, which is in point of fact an interactive book. At the same time computing capacities are actually not used. Only the few, devoted in addresses, passwords, access methods, have opportunity to connect to high-performance supercomputers and to solve their problems. Thus to speak about any intuitive user interface - it is useless: each server lives under its own laws.

 Besides, supercomputers most often also don't suspect about existence of each other, and therefore ones of them are busy with work while others stand for hours in a mode on duty though could, having assumed part of tasks, to promote the fastest satisfaction of users. In such situation appearance of some intellectual system of use of computing capacities of  Internet is just necessary. But what it has to be? What main criteria of its construction? How the user  will work with it? Answers to these questions can't be unambiguous. Therefore, without applying for truth, we will try to state our approach to a solution below. First of all, it should be noted that the offered system can't be monocomputer software. Anyway there will be at least two different parts: client-program and server-program. Probably, both parts will be established on one computer, but it only means that this machine  becomes possible to be used both as the terminal, and as a supercomputer. In principle the creation of software, carrying out functions of a client and a server at the same time is real. However, such approach seems not rational and nevertheless will bring into world a certain symbiosis of the client and the server under one cover. As a result, the first criterion of creation of intellectual system of use of computing capacities is formulated: work on technology client server. The main objective of the client will be interaction with the user, clarification of his requirements and desires, and then search in Internet resources for the solution of an objective. With this the client-program has to have a possibility of interaction with the most popular WWW browsers (for example, Netscape Navigator, Mosaic). It is essentially important because the importance to make some calculations on the remote computer, shouldn't force the user to refusal of usual means of navigation in Internet. In our opinion, only uniformity of network operations is capable to make work in Internet easy and available. The main duties of the program server will become the interpretation of inquiry of the client, assessment of labor input of work, and then its performance or readdressing to the neighbor, which is capable to cope with a task in shorter period. Also, probably, the server will have to contain information system to all available computing capacities of Internet that at the time of place definition for order placement, the client could address to the next supercomputer and receive all interesting information. The second criterion of creation of described system consists, from our point of view, in application of object-oriented approach. We won't propagate about advantages of this approach,we will tell only that object-oriented systems became de facto standard in the field of projecting of difficult systems. How this approach can be applied in our concrete case it is possible to tell some general words only. First, exchange of information between client and server parts will represent an exchange of objects. The question of structure of such objects probably remains opened for a long time, but that the client will begin to generate objects inquiries, and the server to work with them and then to return objects answers - undoubtedly. Secondly, most probably, in creation of system will be used some specialized language for writing of network applications of Internet, for example Java. And it is completely object-oriented. Therefore both constructions of client-program and server-program will be completely object-oriented. As the third criterion, it is possible to call the requirement of the greatest possible simplicity of the user interface. That is the interface has to be intuitive and if to expect studying from the user, it will be minimum and chiefly not labour-intensive. Is it such interface? First of all, it has to have a double essence, that is provide to the user two modes of work with system. The first interface - hidden, actually anything except control panels visually not representing. Its main objective to provide at the request of the user data processing for local appendices on remote servers. In practice it means that if it is quicker to use more economically computing capacities of Internet, it becomes truth. Let’s  tell, that  the computer of some difficult 3D image is carried out not on PC in days, and on Selicon Graphics in some minutes. The first interface doesn't force the user to study any new ways of data input, interaction with a network, etc. It is extremely simple in development, though imposes certain restrictions on the formulation of tasks as those are densely connected with concrete supplements. The second interface has to serve for expansion of opportunities of use of computing capacities of Internet. With its help the user will be able to request carrying out as much as sophisticated calculations and to receive the answer in a convenient form.