Every time when I see a new thing, I try to note it somewhere and come back later. The "later" is driven by how many times I will meet that thing. More often I see it in the books or conferences, earlier I deep delve into it. And that's the story of this post about reservoir sampling algorithm I met twice during last month.
