Programmers will love and value flexibility and mightiness in languages they use, not to speak of Ruby programmers who know more than well that a powerful language can do a lot more with much less effort.

However, there are many situations where less is more. Good constraints and restrictions can make some sorts of things work better, where human errors must be avoided, security concern minimized, or future extensibility reserved.

In this talk will be discussed with real examples what are the trade-offs an implementer should consider, to help the audience make the right choice in their API and product feature design.