and you really think that a student with no industrial experience and no access to statistics from the industrial world, can conduct such research seriously?
--guy shlomo bauer wrote:
HI, As a former professor teaching software engineering, I was bit surprised by your posting -- perhaps I misunderstood your intent. Although software engineering in the large is more about process than code that's not always the case. For example, software systems benefit from code refactoring. An example of refactoring is finding sequences of code that are repeated in a variety of places and replacing them with a function call. The resulting code has the same "meaning" but a different text -- the refactored code is easier to understand, etc. Writing a compiler inandofitself is not a software engineering project. A good project for you might be to look at a tool like valgrind. Consider how such tool can be incorporated in the software development life-cycle. Having done so, you might then try to find a taxonomy of defects (NIST in america published) by frequency, severity, etc. The interesting question then is what set of tools would be useful in helping uncover defects likely to be encountered by customers as well as ones that are catastrophic. If you really want to write code. why not do a comparative study of perl and haskell for a variety of scripting. Why these two? Because haskell was a big win for perl 6 (I'll leave it to you to find out why). from a software engineering perspective, language selection should be based on something more than, "all our code is in perl." Shlomo _______________________________________________ Linux-il mailing list Linux-il@cs.huji.ac.il http://mailman.cs.huji.ac.il/mailman/listinfo/linux-il
_______________________________________________ Linux-il mailing list Linux-il@cs.huji.ac.il http://mailman.cs.huji.ac.il/mailman/listinfo/linux-il
