>I have taught for 4 years and I have had around 600 students of
>computer science.
>Only 1-2 could write code without many errors.
>The fact that little developers can write code with few errors
>it is a fact, and this cannot be eliminated.

Do not mean to offend, but I feel I should stand up for the programmer here.

It has been my observation that programmers have to adhere to schedules. To keep their job they have to write code which just just barely conforms to requirements. If the code has bugs, if it fails QA, then it might receive an allocation of time to fix the bug ... time which was not available originally. Or the code may pass the tests, which are based on the specifications, to which the programmer wrote the code. If this the case, then it is the writer of the specification who is at fault, not the programmer.

Often the programmer would love to make it work BETTER, but just has time to get it DONE.

Interestingly enough, I had a professor in an assembly language course who lectured to the class that if condition A existed and operation B occurred, then situation D would happen. I had worked with the processor before and knew the situation D could not happen. I wrote to Rockwell, the chip manufacturer, and asked for clarification. When I received a letter back from an engineer there -- saying that condition D could never happen -- I attempted to show the letter to the professor, who refused to even look at it. He said my grade was based on how well I learned what he taught. So the problem is not always that the students are imperfect, GIGO. In the U.S. we need better instructors with better pay who must compete for their jobs, not those who cannot handle a job in the private sector.

Regards

Art