Inspiration

Well, we went through the docs and the nightpaper, tried to understand what midnight is all about tried to think of some great ideas watched the workshops etc. We went accross the community board with ideas, saw the face recognition, the kyc repos, realised you can be sure someone you meet is exactly that person, we thought of all the posts we saw on X about how, the first succesful dapp wont be shouting out its on chain, it would just work , we thought for a while and decided hey lets build a dating app (we're hoping this one stands out and is pretty unique, we went ahead actually eliminated ideas from LLMs) I mean current dating apps have several problems, they're obviously not keeping your data safe, having preferences like (looking for a man with a job) might come off gold digger like but has real reasons, you should be able to put out preferences without everyone judging you, the nightpaper made us think yeah wow this could finally be possible now

Our Demo link: https://www.loom.com/share/56efc9c87f534eb4a2f55879aabf0174

What it does

For now, we have a beautiful home page, user onboarding, while onboarding we run our main smart contract and that handles taking in age, keeping it private and you match with ranges using some smart zk math, we ask for location, once again some smart zk math then takes you to a dasbhoard with users you can match with (kinda had to fake data cuz well no users lol, its a proof of concept), the idea you'll see matches there and you send match requests to their wallet addresses

How we built it

Well, the first 30 hours or so, we took the honest approach , read all the docs , looked through a ton of examples, tried to find past hackathon winners, used midnight academy, went through all the scaffolding, had something basic set up my teamate made a beautiful landing page we followed up docs got the connect wallet logic working and some other stuff, but then it came to actual smart contract making and well, it was like nothing I've ever done before ( I have some experience with solidity and rust (solana) dapp making ) the idea of a proof server, docs ngl need A LOT OF IMPROVEMENT, examples outdated, somehow hackathon winners have no .compact contracts on their code? what even so much confusion, the fact that you make the circuits and the zk proof tf? I thought there would be nice library support, you pull a library for location verification etc ) so yeah I started studying zk proofs, watched a cyrfrind updraft course on them but well the clock started ticking on us I had to get moving fast, this is where our use of AI started which I'm not really proud of, https://web3fast.dev/ we tried to make use of this tool along with that we used context7 fed it to copilot we have a few contracts made, some logic there but honestly, some parts are not true zk our idea was ambitious and went super complex fast, we made a comprimise some of the contracts yeah they work, do what they promise but some stuff (faked with typescript lmao) not proud of it but well its a compromise we were too close to the deadline and I think it half works and is a good proof of concept of what could be (maybe even within 10 weeks of us being in that nice accelerator program, dedicating ourselves too midnight)

Challenges we ran into

Well, I yapped a lot about the challenges already but here is a tldr, examples outdated, different from each other some not even using .compact , docs went from hello world to so here's our api , merkletrees and other stuff I'd never seen before

Accomplishments that we're proud of

Proud of how much information we were able to take in and learn in such a short period of time, somewhat string stuff together but honestly there's more stuff I'm not proud of like not being able to implement true beautiful well audited smart contracts.

What we learned

Well, I learnt a lot from the docs and midnight academy, learnt how private actual blockchains really are, learnt so much about zk proofs, uk the standard alice bob entering a tunnel with a door and a secret passkey, overall it was fun its just sad that we got overambitious but we did our best and have some contracts and circuits we're proud to have and understand

What's next for dApp

What's next is mainly upskilling ourselves, build a real understanding of zk proofs and the compact language, get the features actually working and build something beautiful

Built With

  • compact
  • copilot
  • midnight
  • next
Share this project:

Updates