GitHub Nomic Game #3 | Hacker News Reader

GitHub Nomic Game #3

jefftk.com

40 pointsluu14 comments

I admit I had never heard of the game before, I'm trying to reason about how the game would work in person beyond 'give me infinity points'. Has anyone got any links to actual playthroughs?

https://agoranomic.org is one of the biggest and longest lasting games of Nomic.

If you proposed a rule change that gave you infinite points, the other players would vote it down, because it's easy to see that that rule rewards you at their expense.

You can land self serving rules though, you've just got to be sneakier, and slip them in as loopholes or the like.

My god, this all seems so complicated.

https://www.mail-archive.com/agora-official@agoranomic.org/m...

To be fair, we're in the middle of a big mess right now (we discovered a bug in a definition used by many rules, which means that many things don't work; we're working on a retroactive fix). That sort of mess happens every year or so.

(But then, Agora still is really complicated.)

The list of current rules is both impressive, and very intimidating https://agoranomic.org/ruleset/flr.txt

There are many nomic games which involve a starting rule set and a subset of the rules that are mutable. Those are the most playable ones in my estimation.

The initial ruleset requires unanimous agreement for each rule change. This ensures that anyone who loses has, at some point, allowed it to happen (unlike, say, a majority vote, which could victimise a minority of players)

From https://en.wikipedia.org/wiki/Nomic

"While the victory condition in Suber's initial ruleset is the accumulation of 100 points by the roll of dice, he once said that "this rule is deliberately boring so that players will quickly amend it to please themselves".[1] Players can change the rules to such a degree that points can become irrelevant in favor of a true currency, or make victory an unimportant concern. Any rule in the game, including the rules specifying the criteria for winning and even the rule that rules must be obeyed, can be changed. Any loophole in the ruleset, however, may allow the first player to discover it the chance to pull a "scam" and modify the rules to win the game. Complicating this process is the fact that Suber's initial ruleset allows for the appointment of judges to preside over issues of rule interpretation."

So not only would 'give me infinity points' presumably not win a unanimous vote; but the whole idea of "getting points" is designed to be gotten rid of in favour of something more interesting.

I was one of the players in this super fun experiment, and one thing I didn't expect was how quickly things seemed to move.

I have no idea how other online Nomic games work in that regard; someone posted a link to a mailing-list version, and it seems like it would be nearly a full-time job keeping track of everything.

If you are speaking of Agora, the traffic is pretty low nowadays. The main difficulty is getting up to speed with the huge backlog of "jurisprudence" accumulated over the decade long game...

Are there any games running out there that are explicitly timeboxed?

Nomic sort of reminds me of Mao, except rule changes are done by the winner of each round and nobody knows what the rules are (obviously this is rife for abuse, so there's usually a gentleman's agreement to not break the rules too badly).

Why did he push the branches one at a time?

  for i in {1000..2000}; do
    git checkout testing-$i
    git push
  done

can be rewritten as

  git push origin testing-{1000..2000}

which also avoids the useless checkouts.

I love it! I think I might pitch this as something to do with coworkers under the guise of learning more about CI tools.

We already use Jenkins, but the SRE team manages it, so us programmers don't get to play with it too much. This could be a great excuse for tinkering with the more integrated offerings from Github/Bitbucket/Gitlab.

My mistake. It appears that Github still doesn't have any first-party CI offering. I must have confused it with the other two.

Reply on news.ycombinator.com