4 ms·

Or even better, how about empirical research to back up the claim that functional programming is "better"? Does the use of functional language result in fewer b

Or even better, how about empirical research to back up the claim that functional programming is "better"? Does the use of functional language result in fewer bugs at less cost? Compared to what (C, Java, Ada, Lisp)? I don't have an opinion on it but I would like to see the bold claims backed up by actual evidence instead of assertions.

jaredklewis 9y ago
It would be nice, but is a rather tall order. Total cost of maintenance and development of software is not usually public, often unknown even to the developing company, and not actually a very useful metric without context of how much value was generated by that cost. And again that generated value is generally not public and usually unknown even to the companies themselves. Another approach is to run experiments but they are universally terrible. Their methods have so little relation to the way real software is developed, that it is hard to take any conclusions seriously. I would also like some hard data, but don't hold your breath. Technology decisions will probably need to made based on other factors for the foreseeable future.
- BaronSamedi 9y ago
  I agree that hard data is difficult to come by. Without it though, people should not be claiming that X style of programming is better than Y. As much as I like the functional style, it could be that Pascal-style structured programming is actually the best. Who knows? There was a recent post here on HN about empirical language design and I thought that was a refreshing change. I'd like to see a change in software culture towards an evidence based approach. It's long overdue. Alan Kay called the software world "pop culture" and I think he is right. Software is more fashion driven than Vogue magazine. Academic research could be a vehicle for a more empirical approach. Research in the hard sciences embraces empiricism. Sadly, the academic research in software doesn't, generally, adhere to the same level of rigor.

jaredklewis 9y ago

It would be nice, but is a rather tall order. Total cost of maintenance and development of software is not usually public, often unknown even to the developing company, and not actually a very useful metric without context of how much value was generated by that cost. And again that generated value is generally not public and usually unknown even to the companies themselves. Another approach is to run experiments but they are universally terrible. Their methods have so little relation to the way real software is developed, that it is hard to take any conclusions seriously. I would also like some hard data, but don't hold your breath. Technology decisions will probably need to made based on other factors for the foreseeable future.

BaronSamedi 9y ago
I agree that hard data is difficult to come by. Without it though, people should not be claiming that X style of programming is better than Y. As much as I like the functional style, it could be that Pascal-style structured programming is actually the best. Who knows? There was a recent post here on HN about empirical language design and I thought that was a refreshing change. I'd like to see a change in software culture towards an evidence based approach. It's long overdue. Alan Kay called the software world "pop culture" and I think he is right. Software is more fashion driven than Vogue magazine. Academic research could be a vehicle for a more empirical approach. Research in the hard sciences embraces empiricism. Sadly, the academic research in software doesn't, generally, adhere to the same level of rigor.