I agree with Michael. Software Engineering is in a deplorable state for these reasons. It has a real impact in the economy because maintenance costs and failure rates are unreasonably high. It compensates by providing work for lower skilled workers prepared to work hard.
It's tolerated by some (start-up) corporations only because a small team of motivated, productive and hard working bumblers can build stunning and unreasonable corporate financial value. The costs are born by other corporations (and the public) because they know, or can get, nothing better. We are in a growth phase of new technology usage. It will continue for awhile but it won't go on forever. At some point in the not too distant future Software Engineers will need the skills of Mathematicians (and no, they do not have those skills now). They will need these skills because it is ultimately the only way to manage the increasing complexity. The fact of the matter is that Ph.D.s often do write less code, and that is often because less is better. It is also because they typically spend more time thinking about the problem in order that they can write less code. If they have the mathematical skills I allude to then the code may never need maintenance (because you will be able to prove that it has the desired properties). But when placed in an environment in which there is an urgency to patch failures or there is a deadline that does not allow time for thought they often can't compete. The hacker will be needed as long as we write software this way. But it is the wrong way to write software. I've been both BTW, I began as an industry hacker long before my Ph.D. - and I've run project teams with both. Speaking now as an Academic. I use XML, XSLT and XQuery in my research. I do so primarily because system design and programming with schema awareness is a step in the right direction. With respect, Steven On Oct 12, 2011, at 4:22 PM, Michael Kay wrote: > >>> >>> Few of the cs phds I've interviewed could do ANY of the tasks you quote. >>> None had to pass an exam in making programs that actually worked >>> >>> > I have to say my experience is the opposite. I've worked with a great many > software developers who were good at making things, but lacked the education > to discover the theory of how they ought to be made: they were mechanics > rather than engineers. As a result I've seen a lot of people building things > using home-grown invented techniques that were vastly inferior to the state > of the art available from the research literature. Or doing crazy things like > trying to parse XML with regular expressions. You get something that works > most of the time (if you're lucky), but often costs a lot more and performs a > lot worse than if the designers had had a higher level of professional > education. > > Of course that doesn't mean that everyone with a PhD is a good programmer or > designer, but most of those I have worked with have been. > > Michael Kay > Saxonica > _______________________________________________ > [email protected] > http://x-query.com/mailman/listinfo/talk _______________________________________________ [email protected] http://x-query.com/mailman/listinfo/talk
