| Valery Kolesnyk 2007-12-14, 7:25 pm |
| Designing in the industry, in construction is provided by the theory.
The theory provides reliability of designs in these areas. Until then
the theory will not be created yet, in Software engineering the
demanded level of reliability will not be achieved. In Software
engineering theories still are not present till now.
Sphere of human activity, which possesses the theory (especially the
theory in the mathematical form), has huge advantage. Theoretical
statements, namely proofs of theorems, decisions of problems,
conclusions of complex logic designs, systems of axioms, can be
checked up by many people (professionals). Such check practically
excludes mistakes in the theory.
Other state of affairs has developed in programming. If the program is
checked by means of FORMAL METHODS the amount of the people checking
the program, is rather limited. And though as a result of check the
amount of mistakes decreases, nevertheless, they remain in the
program.
Principles of designing of algorithms should be tested and checked
within the limits of the theory, and algorithms and a program code
should be generated by the generator.
The theory should be fundamental with a high level of abstraction,
generalizations and productive laws.
In Software engineering the paradigm now has affirmed: "Theory is
something useless and is unpromising to spend money for its
development. If someone is going to develop the theory let develops it
somewhere, but after let will develop also something practical and
concrete, and we after that shall estimate result ". Such point of
view is similar to that, as though to Einstein in the beginning of his
scientific career have offered for acknowledgement of ideas stated in
his articles to develop also a demonstration variant of a nuclear
reactor or a certain beta-version of a nuclear bomb.
But fundamental theories are not developed in garages.
Certainly, there is a big risk of that the money enclosed in
development of the theory, will be spent vainly. The theory can appear
unproductive. But, despite of risk, it is necessary to search and
develop the theory until there will be a hit in the purpose.
It is necessary to distract from the monitor and the keyboard, from
programming and to sit down with the handle for a table for the
analysis and that generalization that is already known in
programming.
Without the theory there will be no reliability.
More detailed information on SITE http://WWW.k-v-g.pochta.ru
Yours faithfully
Valery Kolesnyk
Ukraine
|