Discreet: powerful eDSL for bitcoin contracts (PoC, in development), + oracles

[dk14]

Docs and tryout: https://dk14.github.io/mega-peers/docs/#/dsl

It is part of a larger initiative (truly decentralized oracles MVP, use PoW to certify facts): https://dk14.github.io/mega-peers/docs/front.html

Sources and donations: https://github.com/dk14/mega-peers/

----

The key feature, aside from being able to support any contract on BTC, is that this DSL does a type-check to prevent "perfect hedges", that result in funds unintentionally locked (for a duration) in contracts... for no reason. It removes illusion of profiting regardless of outcome - no free lunch.


-----

Related. I conceptually developed better use for PoW for fact certification. If you like Magic Peers and Discreet - feel free to check this concept out: https://www.upwork.com/services/product/development-it-yaqui-p2p-search-engine-standalone-web-2007000568819406128

-----

Disclaimer: I developed projects above initially for Cardano and Wolfram (as a short-term gig). I am not a fan of neither, so I figured better to continue them with Bitcoin community, since everything is open-source and neither Cardano nor Wolfram have resources/competence to lift it off. Plus, I convinced them to target BTC too (so everything is catered to BTC in there), since what they wanted wasn't really appropriate: their own services are low-quality and unreliable, politely speaking. They invest into ads much more than tech.

[BattleDog]

Interesting work, but I'd separate the useful engineering idea from the marketing gravity around it.

The useful part is the eDSL angle. Describing conditional payouts in TypeScript, enumerating outcomes, checking collateral sufficiency, and refusing obvious "perfect hedge" nonsense is a sane direction. A lot of contract systems let users lock stupid amounts of capital because the tool has no opinion. Having the tool say "no, this branch structure is economically redundant" is actually valuable. That is the kind of boring correctness feature people only appreciate after they've had funds parked in a contract doing absolutely nothing except aging like cheese.

Where I get more cautious is the oracle identity part. PoW can make Sybil identities more expensive, sure, but it does not magically make an oracle honest. It proves burned work, not truth. Reputation-after-the-fact helps only if the oracle cares about future business more than the one bad settlement in front of it. That is fine for some markets and useless for others. The old problem remains: if the external fact is dirty, late, ambiguous, censored, or intentionally lied about, your beautiful contract tree still eats garbage for breakfast.

[Orange89]

honestly this discreet dsl thing is pretty cool
i like that it stops u from making contracts that lock funds for no reason like how many times have we seen money just stucks because of these code had a weird edge case hmm having the tool catch that before u deploy is actually useful definitely
the typescript style syntax is nice too makes it easier for regular devs to jump in without learning some crazy new language from scratch
on the oracle part with PoW  i get the idea of making fake data expensive but yeah it doesnt really prove the data is true just that someone spent electricity still feels like theres a gap there between like hard to spam & actually correct
also curious if this works well with multisig or lightning like can u build a discreet contract that interacts with a lightning channel or is it just for onchain stuff right now
one small thing  the docs are good but maybe we can add a few more real World examples like show a simple escrow or bet contract end to end so people can see how it all fits together
anyway props for building something that focuss on safety instead of just hype bitcoin needs more tools that help the people not shoot themselvs in the foot gonna play around with the playground & see what i can break lol