Software development is a delicate process that has as a final output a software, the question is, how dependable is such output. How do we know if it is going to work or not and why it wouldn’t work.
I worked for several Mozambican software development companies and unfortunatelly had to deal with software failures. By software failure I mean, it was not solving the problem, it was only causing more.
The truth is the timings are never going to be on our side, but sometimes the timings are just impossible to obey while following every software engineer step, this is where the problems begin. When the timings are impossible following the less risky path, we shall not try using the most risky. It is not viable.
There is a phase on software development where the viability of the software is tested. This phase says if it is possible to build the software, if it is economically viable and if it solves the problem. It happens before the project phase. Most of the times, Mozambican software development companies go by the risk path and most of the times they fall int the same problems : maintenance and hight cost.
They believe that going strict to the point (coding) is faster because they do not waste time with documentation stuff. The pain after all is miserable. Most of them get dependent of the software developers because nobody else knows how it works. There is no software project, there are only source codes. This is where they fail. How is a software project going to move forward if alter all it is not a project!This is why people sketch their houses before they even build. This is what i MEANT by “risky path”.
They never know how much money they will have to invest on a project , they never know what are the critical points of the project, they never know even how is it going to look like and when they finally stop coding, it is not what was expected. THE SOFTWARE FAILS.
All software that is candidate to success shall follow the software development phases : conception, requisites engineering (interviews, questionaries, etc) + system proposal + viability check by stakeholders, project (database modelling, GUI, hardware specs, more documentation, approval by stakeholders), coding, testing and evolution.
The stakeholders intervention in the current mozambican software development scenario is BLIND. They do not have informations to evaluate the project. They evaluate the final output, when they can’t do so much, after all the loss. Most of the times there is no test phase. They believe in the developers unity tests. Once the coding ends the software goes to production.
THIS IS WHY IT IS ALWAYS THE DEVELOPER FALL. The answer is, yes, it is predictable.
Comentários Recentes