Hacker News Reader

Comment by nohat | Hacker News Reader

It is interesting how long it has taken for forensics (even just some small aspects) to be approached in a scientific way. It is very disappointing when you consider how many people's lives and livelyhoods depend on forensics. I wonder what other systems depend on anecdotes and apprenticeships for expertise, that really ought to be approached scientifically.

View on HN

> It is interesting how long it has taken for forensics (even just some small aspects) to be approached in a scientific way

I know someone who contracts out to forensics labs and police depts. The situation is basically a mess. "Experts" are selling "expertise" to people who are not familiar with the domain. So it just comes down to convincing the prosecutor, attorney, judge, a member of a jury, that you are an "expert". That can be done by clever sales techniques: look professional, know how to use lots of buzz words, and stuff like that.

Then the more a certain "expert" is invited to testify or to examine evidence they accrue "credibility" regardless if they ended up making shit up on the spot or not. This "expert" has been invited by appear at 10 trials, so they must be good. Or "look he worked with 3 police departments, he must be good".

The fundamental problem is those who are not experts in a domain need to find expert, so that opens a possibility for "clever salesmanship" therefore my friend's business...

Software Engineering.

I am not aware of an attempt to scientifically approach software engineering generally (though it seems like the type of thing that would have been tried). I have seen some pretty scientific approaches to performance and efficiency. What do you think should be done? Perhaps more peer review. Blinded testing of code complexity and development difficulty for different code systems?

When people say they think software engineering should be approached more scientifically, usually what they mean is that they want a scientifically-verified process that repeatably produces optimal results. Unfortunately, what science has been done has generally shown that good people produce good software, bad people produce bad software, and that it doesn't seem to much matter what process you lay on top of them. (To the extent that sounds tautological, well, I'm summarizing here.) It is broadly accurate to say that what scientific evidence there is about software engineering leads to the conclusion that science isn't going to give us that science-approved unified single process that people are asking for. So we'll continue to hear this complaint for at least the next several decades.

(Science can nibble around the edges of the problem of software engineering, but even what results we have strike me as likely to suffer from the usual problems of taking small isolated samples from an n-dimensional space and then trying to extrapolate. For one thing, almost every study you've ever heard of that establishes some "fact" about software engineering was done on students. Scientifically speaking, there's no particular reason to expect such results to translate to professionals in any particular manner.)

It would probably look a lot like an architect's license.

Institute for Software Research: http://www.isri.cmu.edu/

Software Engineering Institute: http://www.sei.cmu.edu/

Max Planck Institute for Software Systems: http://www.mpi-sws.org/

UW CSE Software Engineering Research: http://www.cs.washington.edu/research/software.intro.html

Reply on news.ycombinator.com