Flame: confidential transactions with Bitcoin proof-of-burn consensus
Oleg Andreev
The network architecture combines proposals by Bitcoin developers and a few ideas from other networks, as well as my earlier work, which some of you may know from Chain's TxVM (2017) and Interstellar's ZkVM (2019). Flame provides transaction confidentiality with a programmable constraint system based on Bulletproofs and Ristretto, Taproot-based predicates, Utreexo storage for unspent outputs, and actors with leased storage. I've aimed for a balanced, practical solution for decentralized applications and privacy, with a security model as close to Bitcoin's as possible, while having a purely market-driven adoption strategy that does not require changes to Bitcoin's consensus rules.
The overview is available at:
https://runflame.org/flame.pdf
The work-in-progress implementation:
https://github.com/runflame/flame-lib
The project is still under development. Mainnet launch is planned for January 2027.
Oleg.
waxwing/ AdamISZ
Hi Oleg,
Interesting work! Personally I like this general class of ideas. If we think about the simplest form of the idea: to enter the sidechain, you burn bitcoin, I believe its first proponent is Ruben Somsen with 'spacechains' [1]. I think you should compare your design with that? At the time he proposed it I was less positive; the obvious question/problem is that without the reverse path the 'coin' can't truly be called bitcoin, it's a different thing. And I also thought, it's hard to convince people a one way transfer is worth it (yeah, that's really vague; who knows; it's more just intuition, less fact). More recently, I put together a PoC of what I called "hodlchain" [2] which, apart from fancy theorizing about time value of money aspects, is basically just 'spacechains but with CLTV/fidelity bond style time locks instead of burn'. I liked this version because I saw it as an unlock of dormant value: the hodler gets rewarded for the positive externality of their behavior without having to change it.
With regard to your minting and supply as per your Section 3: yes I wrestled with that a bit in 'hodlchain' too. The reason I didn't like your minting schedule: fixed per epoch minting --- is that it doesn't feel like the fairest option for the ordinary user: they're forced into an auction/market that they have to make a judgement about whether they're going to get good value or not. But in holdchain I had another 'TVOM' variable to play with (r) due to using locking. With burn, I guess your choice must be right: fixed minting lets the other side (burn) vary according to market's assessed value of flame; if you had a fixed rate of flame:btc in minting then if the market doesn't like it, the system doesn't work (your consensus relies on the minting).
Also, it's unfair of me to compare: you're actually trying to build *consensus* from the burns; I wasn't bothering with that. That makes your design intuitively seem like it must be the correct one: proof of sacrifice of value ~= proof of work. I guess it's 1-layer-down PoW in the sense of, it has PoW security assuming bitcoin's value and consistency, so, assuming Bitcoin's PoW security.
I won't comment on the detailed design of the flame-sidechain itself. It looks pretty interesting :) I think this 'it's not bitcoin' thing is the main thing that puts people off, though a counterpoint is, systems can get quite some interest if they have a speculative element from mining for new/minted coins. The problem with that angle is there are lots of other places to put your bitcoin into another system, speculatively, with all kinds of new features. And it's not hard to get your bitcoin back when you want to via swaps.
Cheers,
AdamISZ/waxwing
[2] https://github.com/AdamISZ/hodlchain
Oleg Andreev OOO
On burning.
We have considered a one-way peg as Ruben proposes (meaning, when X units of BTC are converted one-way into X units of "SideCoin"). The problem is: such conversion erases the information about the risk that earlier adopters of the sidechain take compared to later adopters: if you know that coming later gives you the same reward opportunity, why come in now and bear the higher risk? Then, if you try to tackle that problem with an additional RewardCoin, you face another problem (I really did consider a dual-coin architecture for some time) — how do those assets play together, and how does SideCoin erode the value of RewardCoin (in other words, erode "efficiency")?
If we look at the risk/reward from the angle of network security (as opposed to the individual's financial calculation), the problem is the same. Sacrifice-based consensus, such as Bitcoin's PoW, requires maximum efficiency of the sacrifice: so that every penny of a sacrifice goes into the security metric. If the honest nodes are not the most efficient, but the attacking nodes are, then attackers have a better chance of reverting the chain with less capital than all the honest nodes combined. Making the reward coin weaker as an incentive makes the sacrifice less efficient. For example, by locking the capital temporarily instead of permanently, only a portion of its bitcoin value (the TVOM) is actually contributed, not 100% of the coins.
The difference between RewardCoin and SideCoin is much smaller than one may think because SideCoin would still be a qualitatively different asset with a different and changing market value. Flame simply acknowledges that and offers a fixed supply for whatever the market wishes to burn. In the long run, ≈99.99% of flames are produced out of X% of bitcoins.
On utility.
"It's not bitcoin" is a fair concern, but so is the security of all the other chains where bitcoins are supposedly bridged. To be "the bitcoin", the functionality of interest must be a part of Bitcoin's ossifying consensus rules. With every year, that possibility fades away similarly to how we'd see an astronaut falling into a black hole: never quite disappearing, but surely on the way out :) The problem with soft/hard fork consensus upgrades is that everyone needs to "buy in" simultaneously and believe in the upgrade before it is tested out in real life. Flame is adopted by the market: everyone decides when and what portion of their bitcoin to "upgrade".
Also, there are places that your bitcoins may "visit", but very few where they want to "live". Flame hopefully gives you security model much closer to Bitcoin's and, correspondingly, an asset that's by its nature is much closer to bitcoins than any altcoin.
Cheers,
Oleg.
> You received this message because you are subscribed to a topic in the Google Groups "Bitcoin Development Mailing List" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/topic/bitcoindev/Ov92xNLvm8w/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to bitcoindev+...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/bitcoindev/ccb16c4b-ede1-41dc-acf1-d4c286181e7dn%40googlegroups.com.