jeffbradberry

Yes.

Desktop. I have 3 old machines, just decommissioned at the beginning of the year, that I've been kicking around the idea of making into a cluster.

I have some old machines that take DDR3. What do you have?

What I _really_ want, now that I have the Faber-Castell, is a Curta.

https://en.wikipedia.org/wiki/Curta

I have four now, and even go out of my way to use them a bit.

My favorites from my collection are a little Dietzgen No 1773, which is surprisingly featureful for something that can fit in a pocket, and a Faber-Castell Novo-Biplex 2/83N, which is kind of one of the big guns for slide rules.

That is not accurate -- the Andromeda Galaxy is over 3 degrees wide. About 6 full moons.

This seems like a cool concept, but it's broken for me.

Firefox 113.0.2 shows the orbits and background, but no particles as far as I can tell.

Chromium 120.0.6099.129 shows just a black screen plus the widgets, nothing else.

Both on amd64 Debian Linux.

Update: bumping FF to 121.0 did not seem to help.

There is a fixed probability of matching all numbers but the value of doing so varies, so sometimes the EV can be positive.

Here it is as a pdf: https://rifters.com/real/shorts/VampireDomestication.pdf

This should be the page, but it requires a Flash plugin (so doesn't work for me): https://rifters.com/blindsight/vampires.htm

Sounds like https://en.wikipedia.org/wiki/Expedition_%28book%29

It looks like revealjs.

In what appears to be the originating paper for UCT ("Bandit based Monte-Carlo Planning", L. Kocsis and C. Szepesvári, 2006),

* The choice of action to be returned is the one "with the highest average observed long-term reward"

* For simplicity, the payout value used in the paper is 1=win and 0=loss, which will result in the agents maximizing their wins. Presumably one could choose other payout values (i.e. points for that player in games that have that concept) to adjust the priority of the agents. The mathematics does not seem to forbid it.

* This paper uses as their first experiment a multi-player game. They state, "...for P-games UCT is modified to a negamax-style: In MIN nodes the negative of estimated action-values is used in the action selection procedures." It is straightforward and more generalizable to games with N > 2 players to simply record values from that player's standpoint in the first place, instead of manipulating it after the fact in this manner.

I hope this clarifies some things.

Excellent. I did want to eventually write something showing how to modify this algorithm to handle games with randomness, and ISMCTS at a glance looks like it does just that. Thanks!

I've updated the diagrams with a set of node values that (hopefully) make sense.

Since the win counts are only incremented for nodes "moved into" by the winning player, UCB does correctly model the opponent's behavior here.

Actually, I think it should be 3/8, since it's supposed to be the opposing player's win count.

These diagrams were supposed to just be an update of the diagrams from the MCTS Wikipedia article, adding the correct backpropagation values, but I see there are more problems than that.

A fair point, but in practice I get fellow engineers trying to quote the sequence numbers at me.

I don't understand this obsession with sequential commit numbers. These may vary from repo to repo depending on your commit/pull history.

-1 (if I in fact had the points to do it), for assuming that changing the metric space obviates calling a dimension a dimension.

I have to mark this one "cannot replicate". I see no sign of this behavior on my account.

I found it impressive, and I've seen content here on HN that is less relevant than this.

Not at all in any of the variants I know of.

There's a great little story by L. Sprague de Camp called "Lest Darkness Fall" that is almost exactly this. An archaeologist from the 1930s gets stuck back in the 6th century A.D., and "invents" the printing press (among other things).

Exactly.

Settlers (and Seafarers, and Cities & Knights) of Catan is a great game. My girlfriend and I also like Ingenious. It's a very simple and quick game (and the only one I've found that's still interesting with two players), yet there is some fairly interesting strategy to it.

This article seems to be claiming that computers and text messaging are responsible for the final decline in handwriting, but I grew up before personal computers were ubiquitous (I was born in 1977) and cursive was already heavily deemphasized in the schools I went through. I don't recall a single teacher who wouldn't accept hand printed work.