If it tells you how many letters are right, and how many are in the right place, it could become Datle.

Challenge: build something worthy of being called 204Date (2048 clone).

Smashing together people to win at dating? Isn't that just Tinder?

Tinder goes in two directions. This would go in 4 directions. Swipe up to send them to your friends, down to your enemies, left to the void, right to you.

You start with 2048 single people, and have to act as matchmaker to pair them off into married couples who each produce at least one child. A generation later, you take the 1024 first-borns of each couple and have to pair them off too, into 512 new pairings.

Repeat that long process until there is just one couple left, whose first-born you then propose to. Having lived long and healthily enough to complete this game, you must have some very desirable genes and/or wealth, and you'll be good friends with basically all this person's ancestors, so there's a strong chance they'll say yes, despite the age difference.

Match fruits until you get a date?

Careful, FruitNinjaDating could get interesting if not messy

SHA256 doesn't work like that. Even single bit differences between inputs should result in very different outputs.

Enjoy ;) https://rsk0315.github.io/playground/passwordle.html

It doesn't. But there are so-called locality sensitive hashes.

https://www.pinecone.io/learn/locality-sensitive-hashing/

Sure, such hashes exist. SHA256 is not one of them.

* Flirdle

> Those games only make at least a bit of sense when everybody enters their crush and only when matching both get notified.

Even then, among heterosexuals they will suffer from the same problem as basically every software platform for romance among straight people: the massive asymmetry in interest between men and women. Men would copy and paste virtually every woman they know into the system as a "crush," so they can find which if any of their female acquaintances might be interested, and then they would decide who to pursue.

> Men would copy and paste virtually every woman they know into the system as a "crush," so they can find which if any of their female acquaintances might be interested, and then they would decide who to pursue.

Smart men would, who know their place in the world. Any of them are good enough for you, really. Take the one that you like the most out of the ones who like you. Is this a problem? If it weren't asymmetric, it would be hopeless. Instead, most men get the choice between one or more interested women, and most women get a response from one or more of the men they were interested in.

It's better if one side isn't as picky as the other. The other options are that only a couple of people match up at all, or everyone getting a response from everyone i.e. no signal at all.

edit: I mean, isn't that Bumble?

I don't follow this, either logically or in practice.

Granted I've never dated as a gay man or as a lesbian, but from what I've observed it seems to me a much smoother and more mutual process than straight dating. I've even heard from bi men about how much easier and less stressful it is to date men than women. AFAICT when both parties are on a level playing field, and there's not a massive asymmetry in power, interest, and investment, there's a lot less grief overall.

> edit: I mean, isn't that Bumble?

Bumble is a valiant attempt to fix this issue but it seems to mostly be a failure. An average man may occasionally get a message (which will usually just say "hi"), but the asymmetry in interest is still there, and it's still mostly men's responsibility to do the real initiating.

Indeed, a similar "secure" crush confession on an MIT facebook group was exploited in this way, by brute forcing every name in the student directory.

You'd want to use asymmetric public key cryptography so that Bob could hash his message with Alice's public key, and it could only be decrypted with Alice's private key.

That still lets Alice find out.

I don't see a clear way of mapping the concept of a zero knowledge proof to the crush problem.

But the best I can come up with is:

1) No one learns that Alice (claims to) have a crush on Bob

2) Unless Bob also claims to have a crush on Alice, in which case.

3) Only Bob learns that Alice has a crush on him.

4) If Bob learns that Alice has a crush on him, Alice learns that Bob has a crush on her.

5) All the above guarantees are symmetric if you swap Bob and Alice.

Seems to me such a system must make it expensive to claim to have a crush on someone. Otherwise Bob can just claim to have a crush on everyone and will find out who crushes on him.

For heaven’s sake just call her!

That lets Bob confess semi-deniably to Alice (if Alice is willing to share her private key, she can still show Bob's confession to Carroll and prove that it's authentic). But it still doesn't do any matching; Alice will know about Bob's feelings whether she returns then or not.

Yup, everyone publishes their public keys. No one’s private keys gets published (fake crush created).

I’m not sure if zero-knowledge proofs exist for love outside of action - repeated acts of commitment over time (to reduces the confidence that someone doesn’t love you).

This could be a fun way to teach a class about the vulnerabilities around hashing. Strong hash != strong protection.

Maybe the second half of the class could be about designing a more secure system. Or use this dilemma to explain public-private key encryption.

Yeah, it is like hashing a phone number to protect its owner's identity. With a small domain known by the attacker hashing isn't really one-directional.

Or just tell your crush, without sharing URLs with everyone

Easy as an adult. But as a kid I feel like we jumped through soooo many hoops to de-risk telling someone we had a crush on them.

I remember those reallllly early games like MASH or even early "compatibility" apps. Very coy.

Yes, I can see how -- as a kid -- sharing an URL of your hashed crush with the world is not going to be awkward at all.

Putting links in your tiktok/whatever profile doesn't seem like very unusual nowadays...

Awkward is the name of the game.

There's nothing wrong with only wanting to reveal a crush if it's somewhat returned. If it isn't, you're just putting them in an uncomfortable position for no good reason.

And this hash url absolutely doesn't solve this problem

I feel like this could be a good implementation.

"Find out if your crush fancies you, just choose them from your contacts:"

...

"Message 'hey there, I like you' sent"

Or even easier, use a SHA-256 reversing service that has all names.

that's because it's not zero knowledge, it's password hashing.

a zero knowledge proof would do something like prove a valid attestation of love, but reveal nothing about who (other than maybe membership in some large enough set).

Bob's crush, though, works at the ice cream store on the corner.

And when everyone is an actual adult, not a manchild or other people without a fully formed prefrontal cortex and thus prone to making ugly, life-changing decisions - for themselves and others - on a dime.

That's most of the demographic for this product. The number of over-25 people, not in a committed relationship, who still haven't figured out how to ask people out, is pretty small.

I dunno. I'm reasonably well adjusted and apparently even decently attractive, and I still have a super hard time with the whole asking people out thing. I've been lucky enough to have people ask me out but otherwise I'm not sure I'd have had much luck dating at all.

This tool doesn't do what I'd want, but a tool that did would be kinda neat :)