Question

I saw some guy defining configurations (like key, some static names, etc.) inside a class names Constants which has const properties.

I know that it is most elegant way to use configuration files for this kind of scenarios as described here but what is the advantages? I can assume that there must a performance benefit even if it is a little.

Answer 1

The advantages are that you can change those value at runtime, with constants you can only change the values at compile time (meaning if you need to change something, you need to recompile and redeploy).

I still use constants and “Constants” classes for internal settings that I know won’t need to be changed at runtime. These save you a little bit of overhead in reading, parsing, and setting up your configuration classes (ConfigurationManager etc). But if there’s even a remote chance you might want to make an adjustment at runtime or after deployment, the performance savings isn’t worth the extra hassle (IMHO).