Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Why Complexity is Different

mystudentvoices.com

Post reply on HN
Print view View original

1–6 of 6 posts

Re: Why Complexity is Different

#3

Post by milesrout »

I admit I was a little surprised that I got to the end and hadn't seen any 'And the reason this is relevant to software is that for your software to be understandable, grokkable, analysable, etc. you need to think about how the details affect the whole' section.

Excellent article nonetheless.

Re: Why Complexity is Different

#4

Post by szemet »

milesrout wrote: post #3

I admit I was a little surprised that I got to the end and hadn't seen any 'And the reason this is relevant to software is that for your software to be understandable, grokkable, analysable, etc. you need to think about how the details affect the whole' section. Excellent article nonetheless.

I come up with one for you:

The clear separation of microstates and macrostates in software, through proper abstractions is a key to good software quality.

Unfortunately the lack of abstractions or the prevalence of leaky abstractions, due to bad design and software management practices (e.g. monkey patching) eventually leads many codebases into its critical point. In this critical state small perturbations of microstates (either at development or at runtime) causes unpredictable behavior changes at macroscale level.

So if you ever have wondered why our software quality is worse than the laws of thermodynamics would justify:

Our best physicists are now working on theories based on renormalization groups to exactly describe the expected distribution of blue screens of deaths: To have much better models of how bad our applications are! ;)

Post reply on HN