HOLIDAY BUILD-OFF!

[tokumei]

It's been a while since we did an ORE build-off, so the staff team has decided to organize one for the holiday season! We have a couple of questions that we would like to ask:

1. What day would you prefer to meet for the build-off (if you plan on participating)? Fill out the poll above, checking all boxes where you are wiling to attend. The most popular option will be selected. Currently, the plan is to run the build-off for ~ 24-48 hours to give enough time for people in all time zones to participate (and to give slow people like me a chance).

2. What themes / topics / rules would you like to see? Leave a post below, and we will begin voting on them later this week.

[PabloDons]

make the theme elytra obstacle courses

[tokumei]

Here's an idea: Build a machine (optionally reprogrammable, bonus points there) capable of calculating a well-known but non-trivial constant (like pi or e) to a given precision.

[Chibill]

I am working on this so I can record this Build off when it happens.

https://www.replaymod.com/forum/thread/1952

[LambdaPI]

A combination of fastest/smallest Bresenham's line drawer.

I like the pi idea tho.

[LordDecapo]

I like the Pi idea. That sounds super fun... could be graded on speed, size, and accuracy.
I like that one a lot actually.

I had an idea, but can't remember it lol. I'll stick to supporting the Pi idea for now.

Also, I CAN do Christmas. (As I personally am passing on holidays this year, screw them) but would prefer another day since I know a lot of people can't get on then.

[PaukkuPalikka]

I like the pi (possibly e) idea too. (I've wanted to do that for a long time)

[Chibill]

I have a head start on that. As I have been figuring out PI stuff since Pi Day

[LambdaPI]

Also, I CAN do Christmas. (As I personally am passing on holidays this year, screw them) but would prefer another day since I know a lot of people can't get on then.

I can basically do only 22nd which is sad so maybe it's possible for people who wont be able to make it to record and build early?

[optimo]

I want to change my vote to the 30th ;p

[Nuuppanaani]

Here's an idea: Build a machine (optionally reprogrammable, bonus points there) capable of calculating a well-known but non-trivial constant (like pi or e) to a given precision.

I like this idea a lot, but it's a really hard one.

But I guess we've had enough of the build-a-basic-cpu challenges.

I'll probably join as I should have a whole bunch of free time on the holidays ^^.

[konsumlamm]

Here's an idea: Build a machine (optionally reprogrammable, bonus points there) capable of calculating a well-known but non-trivial constant (like pi or e) to a given precision.

I like this idea a lot, but it's a really hard one.

I'm not sure what algorithm you mean, but imo it's not really hard. Sure you need a multiplier for most algorithms, but that's still not much harder than a CPU. I like the idea of finding e or π, maybe some other constants like sqrt(2) (easier)?

Another suggestion would be a simple floating point adder, maybe with a custom format for simplicity. Idk about bresenham, but i think its doable too.

[tokumei]

A floating point adder would be great. And yes, I think a square root might be easier to calculate. Maybe we should put off Pi for Pi Day...

[LambdaPI]

I think calculating e would be great.

[konsumlamm]

Does calculating some fraction that is approximately π count?

[tokumei]

Yes, but you might get a lower score for precision. If you can use a formula to generate any number of digits, that would be way cooler than a divider.

[LambdaPI]

We're gonna have to define some rules tho for stuff like e. Considering the fact that you can basically have a huge tape of random 1/0 generators. If you just count the amount of correct numbers then that crappy workaround would win.

[tokumei]

I can't find the method that you're referencing, but if it is a valid way of calculating e, then it should be perfectly legal. Depending on the scoring, you might lose points for not using a general-purpose machine, but as long as you're not using a hardcoded table of digits of e, it's fine.

[LambdaPI]

Create a string of random bits
Compare to e
Get some amount of bits correct
ez
It was a joke.

[konsumlamm]

but saving e in a ROM seems to be a legit algorithm to me

[Koyarno]

@ nonemu; being general purpose means going more in the cpu route. Bonus points for precision, speed, size but also being able to change precision seem very valid.
@ lambda: i wouldn't discourage people from going the creative route - but not cheaty route, and by that i mean using MC mechanic based RNG.

[PabloDons]

Being programmable for other formulas and constants definitely basically means make a cpu. I think it should be a machine who's sole purpose is calculating a constant with no prior knowledge of the constant except it's formula. A machine that keeps calculating more and more precision infinitely deserves the win imo