1 Amateur Rocket Crashes, Another Explodes
...and you wonder why the majority of professional engineering organizations have no wish of being associated with software developers, or having software engineering as a recognized engineering field.
Quite a few states and provinces, as well as countries, require all professional engineers to be licenced to provide a minimum level of competence and to avoid unprofessional conduct.
In software there is the luxury of solving some problems via this trial and error practice since the typical cost of failure is low. At times the cost of simply trying a possible solution without validation is lower than not; then using trial and error is acceptable. Except for the most rudamentary of tasks, this is usually not the case -- in software development or otherwise.
There is a large difference between live flight tests in aeronautics versus pure software test cases. There are people's lives at stake in the case of flight tests, regardless of whether the vehicle is manned or not. It is not acceptable to have such a cavelier attitude towards developing systems that can put lives at risk.
Taking the attitude that development is a simple cycle of repetitive tests with incremental improvements is flippant at best. There are costs associated with each failure, and engineers have a professional as well as moral obligation to keep these costs at a minimum.
Our ability to create aeronautic vehicles has improved beyond simply having to try an idea to see whether it works. Writing a off catastrophic rocket failure as just another test cycle is reckless at best, incompetent and a malice act at worst.