Re: [PATCH] hamt: New module.

[Bruno Haible]

Hi Marc,

> thank you very much for all the helpful feedback. I have incorporated your
> suggestions, and you can find an updated diff below.

This is definitely better.

> +Lookup of a (key, value) pair given the key is an O(1) operation,

I forgot to say "on average" here. (It's not worst-case O(1).)

> On the other hand, a hash table is
> +simpler so that its performance may be better when persistence is not
> +needed.

The term "persistence" has not been explained at this point.

> +HAMTs are well-suited to a persistent data structure, which means that
> +each updating operation (like inserting, replacing, or removing an
> +element) returns a new HAMT while leaving the original one intact.

Here comes the definition of persistence. Do you want to move it upwards,
or is the talk about persistence redundant, now that we explain the
use-case in part 2?

Feel free to push. It's your module.

Bruno