Hacker News Reader

Comment by melvinroest | Hacker News Reader

I wish more explanations would go like this: iteratively. It's much easier for me to see where I suddenly start to struggle. The author also used language that feels more familiar to me, which helped a lot.

Unfortunately, a lot of writers just dump the information and use terms from different fields. And I get that, but it makes me slower to understand it.

View on HN

I think it's interesting that Lamport chose this style for Paxos Made Simple[0] Which iteratively goes through a process of proposing a solution for a problem (proposal 0: accept the first value the acceptor comes across) and then reveals the problem with it, and amends the proposal to fix the problem

[0] https://lamport.azurewebsites.net/pubs/paxos-simple.pdf

This Kerberos explanation does something similar at a high level through dialogue. Always did like it.

https://web.mit.edu/kerberos/www/dialogue.html

At least for paxos this is more or less the standard method of explaining it, see: https://youtu.be/JEpsBg0AO6o

In paxos's case the logic is very, very simple - it's just the reasons for the logic are difficult to sort out. So teaching it iteratively works very well, because you show how simpler versions fail.

> So teaching it iteratively works very well, because you show how simpler versions fail.

This is, in general, a very good approach for teaching (although time consuming).

The key thing about it being iterative, I think, is that it builds up the justification for seeing that this is an effective and reasonable solution, one step at a time.

Yeah it's sort of a microcosm of the general path of human learning: build a better model, shoot some holes through it, build a better model. I don't think it's any accident that the socratic method has stuck with us so long

Ability to write an explanation like that is a sign of a great understanding of a topic. If you can write something like that - you've achieved such understanding.

Reply on news.ycombinator.com