b52_

I'd argue that one billion is fairly large for what most people work on. But yes, point taken.

What do you mean? Modulo is not a perfect hash function... What if your hash table had size 11 and you hash two keys of 22 and 33?

I also don't understand your first point. We can run n^2 algorithms on massive inputs given its just a polynomial. Are you thinking of 2^n perhaps?

Hey ot,

We worked on improvements to this following our reproducibility study including a synchronized iterative version and alternative estimation functions.

You can find the paper here if you're still interested!

https://jmmackenzie.io/publication/tkde22/

Where do they claim to be best in class? I don't seem to see that anywhere.

Nobody cares if matplotlib is outdated. It served its purpose here, which was to show how plotting data can lead to further understanding data. You don't need a fully featured tool to do so.

> Each one of these needs to be scored (well, sorta - there are various tricks you can use to avoid scoring some docs, which again I'm not at liberty to discuss)

You mean like Wand or BMW?

I completely agree with this article, and the parallels I draw from it (and some of the comments here) are extremely close.

But why is it that I read this article, agree with it, yet then continue scrolling HN or FB 2 minutes later? How can I get out of this dopamine cycle?

I suspect they do indeed use such URLs to track which link the user clicks. It's not that they log the IP with the click, but rather, the click allows implicit relevance feedback with respect to the issued query. This can be used for machine learning (to improve ranking, for example).