Case 13-7

Сегодня пообщался в Фейсбуке с русскоязычными айтишниками и опять убедился, что объяснять элементарные вещи -- это совершенно безнадёжное занятие.

Что, собственно говоря, оказало благотворное влияние на мыслительный процесс. Долго взвешивал, стоит ли надеяться на то, что люди будут действовать рационально, а не попытаются продемонстрировать свою экспертизу в вещах, которые банально не понимают.

Сейчас совершенно очевидно (уже не интуитивно, а на уровне логики), что рисковать не стоит. Сама идея очень красивая и даст интересную информацию даже в случае негативного варианта, но, посмотрев на очередной счёт от адвокатов, решил, что гораздо правильнее сокращать сущности.

Теперь надо думать, как это написать настолько формально, чтобы сработало, и настолько однозначно, чтобы предотвратить попытки поразительных интерпретаций.

В дебрях водопада ("Software Enginering" failure)

Аn attempt to explain obvious is one of the most hopeless tasks. Please consider the information below as a private opinion. If you think that it is wrong, I would not try to change your conclusion. I myself have spend more than 10 years and have obtained a lot of quite specific knowledge to understand the source of problems.

The common legend says, that the great people from the previous century knew engineering methods of the true software development; this precious knowledge was common at the end of the 80x but was lost, weakened, and dissolved during following years; this process was caused by bad managers and dumb developers; and we must return the True Science into our processes to restore the Golden Ages.

Unfortunately, this legend is wrong. This "engineering methods" were defective and the failure of their application in the real industry is the logical consequence of their fundamental flaws. Such organizations as the Software Engineering Institute or Rational did not understand either software nor engineering. The rise of popularity of their ideas was based on the wrong promises and the crash of almost all implementation attempts could be predicted beforehand. (There are always some "success stories" by application of any method, any technology and anything other. In most cases they are produced by inability to get the right information or by inability to see catastrophic side effects.)

First of all, this "engineering methods" were not about engineering, because the true engineering is based on thinking. The "software engineering scientists" had tried to describe the software development process by using industrial production models based on repeating of standard conveyer-like operations.

Consequently the statistics that had "scientifically proven" this "engineering methods" were wrong not only because the experiments were wrongly designed -- I would now not describe this aspect further -- but also because this scientists and consultants had tried to measure wrong things. They had shown what they were designed to show but had failed to find the things that are really important.

Errors in software development -- and in engineering in general -- cannot be described as defects which will be placed by each new production stage into more and more complicated modules what makes this errors more and more difficult to find and to correct. The errors in thinking do not remain localized in one place but cause nonlinear effects in the whole system. This means, we do need a different language to speak about them. (Actually, I have just now drawn a draft of a nice graphical illustration but it is to tiring to prepare a presentable version of it. Maybe I find time and show it someday.)

I could agree that the statistics produced by experiments of the "software engineering" science are very impressive. Additionally I could agree that the application of the methods of this "software engineering" had shown positive effects in many cases.

If traders bring to the wild tribe on a lost island not only the glass necklaces but also iron shovels, this creates miracles. However, this is not the solution because sometimes you need an axe, sometimes you need a saw, and the general anesthesia would be better made by using a strange liquid from the glass ampule.