a journal of a researcher

Wednesday, June 25, 2008

Why Computer Science is Dying

Computer science is still a new science compared to physics or chemistry. But if we count the acceleration of development, maybe computer science is not so young. What we are saying computer science is dying is not to say computer science is no more important, just to say it is over the age of blasting. Or more specifically, the job market does not need so many computer scientists, for academic or for industry. Computing theories for von Neumann computer are studied enough. You’d better move to other computer architectures. Computer applications are still penetrating the other sciences. But this is not classic computer science. Many students in computer science actually will become software engineers in companies.

Software engineering is still not matured. People do not have enough experiences yet to guarantee a software system works. Only a very low percent of projects (as low as one digit, to be verified) are financially benefit and satisfy user requirements. You hear many stories that the projects are delayed, over-budget, not meet user requirements, low performance. Many software systems do not bring any benefits. In other engineering domains, this record would be disaster. You do not hear that a construction project has 50% success rate. We can control the other engineering projects well, but not software engineering projects.

2 Comments:

  • "Only a very low percent of projects (as low as one digit, to be verified) are financially benefit and satisfy user requirements. Many software systems do not bring any benefits."

    I can't agree with this statement. Computer has been applied into our daily life, and it is helping us a lot. Credit card, news website, bbs, facebook, blog... Why are you saying that computer is a failure?

    By Blogger Ben, at 9:15 AM  

  • Ben: she is not saying anything was a failure. She is saying that Computer Science is mature, whereas Software Engineering is not.

    The statement she makes is certainly true. Probably less than 10% of all software projects generate some value for the users. Most software is useless or counterproductive.

    By Blogger Daniel Lemire, at 11:42 AM  

Post a Comment

<< Home