The trilogy of Weinberg's technical thought: subverting the perfect Software (4)

================

Satya Interaction Model

The Satya interaction model first breaks down any communication process into four main stages: ingest, define meaning, determine importance, and react.

Ingestion (intake)

During the ingestion phase of the process, a person obtains information from the outside world. Ingestion is not "just like that"; it also includes the process of selection, and we actually have a lot of choices about what we see and hear.

determining meaning (meaning)

The ingested data itself has no meaning until someone gives them some meaning.

Determining Importance (Significance)

The data may give some hint of meaning, but it will never imply its importance. Why is there a step in determining importance? Because if we do not classify and choose according to the importance of information, the perceived world will be a torrent of data that cannot be resisted. Once we have identified the importance, we can give priority to certain ingestion patterns and their meanings, and basically ignore the other patterns. The importance of assigning to a ingest and its meaning may lead us to change the process of ingestion and determination of meaning.

React (Response)

At the stage of reacting, one will conceive the action to be taken.

================

Information ingestion

People tend to respond differently to information from different sources. If you find that others are more willing to accept the message that you are conveying from another person, consider this finding as a test result.

When people turn their attention to other places, they also omit certain information. Without the attention of others, there is no need to provide information, because the other party is unlikely to accept the information. In an environment of accusations, managers tend to think that raising bets regularly (financially, emotionally, or otherwise) improves testing. But if the stakes are too high, you will only hear the accusations, and the fear that it causes will make people deaf. For example, if the manager believes that failure to deliver the product on time may jeopardize his or her work, they will not be open to hearing information that seems to increase the likelihood of delay.

At some point, people stop or reduce their intake of information to a certain extent if they are bombarded with too much information.

Golden Elephant syndrome: A white elephant may be a big and useless thing, but if it is made of gold and spends a lot of money, it will tempt some people to try to use it anyway. But if it is expensive, the person who buys it certainly does not want to be seen as a fool, and the consequence is to become a bigger fool.

================

Determine meaning

If you can't think of at least three possible explanations for the results of your test data, you're not thinking enough.

If you are not sure what a program is expected to do, you can never be sure that it is wrong. Before you give a meaning to a test report, you need to figure out what you expect.

The data itself does not speak, and different thinking determines different meanings.

================

Determining importance

Different people give different importance to the same information.

Importance depends not only on a single flaw and on the nature of an individual, but also on the context. The expected value of a failure is the likelihood of this problem multiplied by the loss caused by the occurrence of the problem. But even if the expected value is used, there are other complexities, such as who is responsible for the loss caused by the failure.

If we pay attention to the emotions, listen carefully, first solve the important things and then solve the unimportant things, we can make the best deal with the data obtained.

================

Make a response

The correct response to defects should be simple: find them; evaluate (figure) them; fix them.

Reassign the power to complete the repair effort, rather than trying to fix a low-priority flaw. In particular, to clear those known important flaws, let testers help developers to identify their hidden failures. Then, pass the test to discover any defects that may exist in the repair.

Many estimates of software projects fail because of the critical steps that are overlooked. Often they ignore "wrap-up" steps such as identifying defects, modifying these defects, and then re-testing and repairing defects in test changes.

The trilogy of Weinberg's technical thought: subverting the perfect Software (4)