a) Decide on the topic. Sounds absurd? By no means. The correctly chosen and formulated theme determines the whole further process of writing the text. The theme is an extremely concise reflection of the subject area within which you write the text. Theme defines the terminology, the degree of detail of definitions and reasoning, references to external information sources, the number and complexity of the "additional material" (formulas, tables, illustrations). It (this is most important!) Determines the whole style of the text. For example, it's difficult to write a thesis in a scientific style with the theme "Content Management System". It is quite another matter when the topic sounds like "Software tool for managing the content and structure of an Internet resource".

b) Determine the volume of the material. If these are theses of the report for 1-2 pages - it will be necessary to "squeeze" the very essence, only the essence and nothing but the essence. If this article is for 5-10 pages, you can already afford a more detailed presentation, more "additional material". If this is a term paper or a thesis (or even a thesis), then a rather complex structure of sections and chapters appears, an analytical part, conclusions, a description of own developments, experimental studies, etc.

c) Read at least some examples of scientific texts. Yes, writing such texts anywhere is not specifically taught, but there are still those who can write them. Take some scientific journal on the profile of your topic, look for materials from relevant conferences, etc. Read it. Penetrate the spirit of such texts, however pathetic it may sound.

d) Repeatedly rework the text. Yes, in due course you will learn to write a scientific text at once, "from the first call". With time. In the meantime, write the text in several stages: first - how it works, and then more and more of it "learning" (correcting the wording and the structure of the text, adding tables and formulas).

e) Use the "by contradiction" method. If you read a fragment of your text and you understand that it would look great in a popular newspaper, it means that this fragment needs to be recycled. You are not targeting an audience of teenagers and "people close to the computer", you are focused on professionals who do not need to be enticed somewhere with expressions such as "delightful speed", "stunning gameplay", "revolutionary decision", etc.

f) Remember that you do NOT write a work of art. Even if you are so delighted with the idea being described that you can not sit still, in a scientific text  from Edubirdie ,the expression, for example, "an implementation that meets a wide range of quality requirements" "a tremendous realization" looks much better than "a tremendous realization" .

g) Try to avoid technical jargon and a clipped translation of those words for which there are adequate analogues in Russian. Even such "established" jargon as "Winchester", "switch", etc. look bad in the scientific text.

h) Do not explain the obvious. You DO NOT write a textbook for a junior school. If you refer to well-known technologies, methodologies, algorithms, methods, etc., just indicate exactly how you applied them. Do not explain their essence, if it is not necessary.

i) Try to write "not too long and not too short sentences". Of course, this is easier said than done, but it should be pursued. Unfortunately, even among the "professional audience" many people do not perceive the long, complex sentences in a few lines. Break them into separate phrases. At the same time, "short proenozenitsa" in 3-5 words, too, look sloppy. Look for the golden mean.

Remember the structure of the narrative. It is as follows: introduction ("approach to the topic"), the formulation of the problem or problem, the necessary theory on the topic, the solution you are proposing and its theoretical justification, practical research and conclusions, conclusion.

l) Analysis requires withdrawal. If you called the chapter (or part of the article) "Analysis (something)", then at the end there must be some conclusion. Otherwise, what was this analysis for?

m) Remember about math. Matapparat and its manifestation in the form of formulas perfectly "learns" any text. Objections that "this can not apply mathematics" are not accepted. Believe me, even the preparation of soup can be described mathematically. Just a couple of examples: navigation on the site - this is the theory of graphs, search - the theory of sets, multithreading and others like it - the theory of queuing.

n) Use graphic material. The algorithm looks much clearer in the form of a so-called. "Block diagrams" (by the way, they are called "algorithm scheme", "system operation scheme" etc.), numerical data look good in the table or in the form of a graph (or diagram).

o) Etc. 🙂 Read more scientific texts, adopt the best ideas from their authors, learn from others and their experience, practice. Show your "feather samples" to the scientific supervisor or more experienced "senior comrades", discuss with them what was written and how to improve it.