Sosaiem is a cryptocurrency designed so that no institution stands between a person and their money. New coins are created only by proof-of-work, at a steady network-wide rate, up to a fixed and final supply of 12,212,010 SOSA. Unusually, mining rewards are split across all active miners in proportion to verifiable work, giving smooth earnings without any pool operator. Every signature uses a NIST post-quantum standard, so the network is secure against quantum computers from its first block. Coins move without fees: transfers are confirmed not by miners but by stake-weighted votes among holders, where a participant’s weight is simply the coins they hold — making forged identities worthless. The result is a single shared ledger secured by two complementary mechanisms, run entirely by its participants, with no privileged party anywhere in the design.
Sosaiem was built around a small set of non-negotiable commitments, and every mechanism below follows from them:
The total supply is fixed at 12,212,010 SOSA and is never exceeded. Coins come into existence
only when a block is mined; there is no other source. Issuance targets approximately
100 SOSA per hour across the entire network, independent of how many miners participate, achieved
by aiming for one block roughly every minute with a per-block reward of about
1.66666667 SOSA. At that rate the last SOSA is struck around fourteen years after genesis.
Once the cap is reached, block rewards fall to zero and mining naturally
winds down, while transfers continue indefinitely by voting. Because transfers never depended on mining
for their fees, the network does not face the usual “end of block reward” security cliff.
A block is valid only when its proof, read as a number, falls below a difficulty target. Miners search by varying a nonce over a compact fixed-size header containing a fingerprint of the block’s contents, so tampering with any transaction invalidates the block no matter how large it is.
That proof is deliberately memory-hard. Each attempt runs scrypt over 8 MB, which has to be allocated and walked at random. Purpose-built mining chips beat ordinary hashing by packing thousands of tiny cores onto one die; give every core its own 8 MB of fast memory and that advantage collapses, because the memory dominates the cost and the chip ends up resembling a plain CPU. Litecoin used the same idea with 128 KB, which proved small enough for chips to catch up. Sosaiem uses 64 times that.
Work-shares are computed the same way, and this matters more than it looks. Had shares stayed cheap while blocks grew expensive, anyone with fast dedicated hardware could flood the network with shares and collect most of the reward without doing the costly work — and honest miners would have been obliged to carry those shares in the blocks they found. Both cost the same, so there is no cheap way in.
Blocks still carry an ordinary SHA-256 name for linking and lookup. Only the proof is expensive, and each node checks a given block’s proof once.
Difficulty adjusts itself. Every eight blocks — one epoch — the network measures how long the recent blocks actually took and nudges the target so the average returns toward one minute. Each measured interval is clamped before it counts, so a single abnormal gap (a laptop asleep for hours, a node paused) can influence at most one bounded adjustment rather than sending difficulty into a spiral. As more miners join, blocks arrive faster, difficulty rises, and the network-wide rate settles back — which is exactly what keeps issuance near 100 SOSA per hour no matter the crowd.
As with any proof-of-work system, individual block times vary widely; only the average is controlled. This randomness is a feature — it is what keeps two miners from repeatedly finding blocks at the same instant.
In most coins, one miner wins each block and takes the whole reward; smooth income requires joining a pool run by a trusted operator. Sosaiem removes the operator. While racing for a block, each miner also finds work-shares: hashes of a simpler, publicly-checkable puzzle bound to the current block height and to that miner’s own address. A share is genuine proof-of-work — it cannot be forged without spending effort, nor stolen, nor replayed on another block.
Miners broadcast their shares. Whoever ultimately wins the block must embed the shares seen and split the reward in proportion to them, and every node re-verifies both the shares and the split before accepting the block. The effect is pool-smooth earnings — every active miner is paid a slice of nearly every block, proportional to the work they proved — with no coordinator, verified by each participant rather than trusted. A winner could in principle omit some shares it saw to enlarge its own slice; nodes include every share they receive, and the fuller defence — a shared share-chain that discards the blocks of miners who withhold, in the manner of P2Pool — is a planned extension.
Ownership and voting are authorised with ML-DSA-65, the module-lattice digital signature standard
published by NIST as FIPS 204 (derived from CRYSTALS-Dilithium). Unlike the elliptic-curve
signatures underpinning most cryptocurrencies, ML-DSA is designed to resist attacks by large-scale quantum
computers. An address is derived as SOSA followed by the first forty hexadecimal characters of
the SHA-256 hash of the public key. The trade-off is size: post-quantum keys and signatures are far larger
than elliptic-curve ones, which Sosaiem accepts as the cost of durable security.
Transfers do not ride inside mined blocks and never wait for one. When a holder signs a transfer, it is gossiped across the network. Each validating node checks it against the shared rules — valid signature, sufficient funds, not already spent — and broadcasts a signed vote. A node’s vote carries weight equal to its coin balance. Once votes representing more than half of the online stake approve a transfer, every node applies it. There is no fee at any step.
Votes are monotonic and self-healing. Because nodes are sometimes momentarily a block apart, a node might first judge a transfer invalid; when its ledger catches up it upgrades its vote from no to yes. A yes is never withdrawn, so confirmations are final and a replayed old vote cannot undo them. This lets a perfectly valid transfer confirm reliably even when the sender spends coins so fresh that some peers have not yet seen them.
A participant’s balance is the sum of two records on one ledger: the mining rewards credited to them by the blockchain, plus the net of the transfers confirmed to and from them by voting. Both records are replicated and independently verified by every node, so all nodes compute identical balances.
Crucially, voting weight is conserved coin. Splitting a balance across a thousand fresh addresses does not create more weight — it only divides the same weight. Creating empty identities is therefore useless: they carry no stake and no vote that counts. This is what secures the voting layer against the Sybil attacks that defeat naive one-identity-one-vote schemes, and it means validators are not authorities — they are calculators whose influence is bounded by the coins they genuinely hold.
Every participant runs a node that both secures the network and serves that participant a complete wallet and block explorer in their own browser — there is no central web service holding funds or keys. Nodes speak over ordinary HTTP and find each other three ways: automatic discovery on a local network, exchange of peer lists so that knowing one reachable node bootstraps a newcomer into the whole network, and an optional published seed list. Each node independently validates every block, share, vote, and signature it receives, so introductions can come from anywhere without requiring trust.
Sosaiem’s guarantees hold for ordinary use. Its design also has defined boundaries, set out here so that operators understand the assumptions they rely on:
These boundaries narrow as the network matures and as more contributors work on it. Reading the source is the surest way to understand exactly how each mechanism behaves.
Sosaiem has no sale, no allocation, and nothing to buy from anyone. The only way in is to run a node: obtain the client, start it, and either mine to help issue and secure coins or receive them from someone who already holds some. The network is its people. Its future — its value, its resilience, its reach — is not something any single party can grant; it is something participants build by showing up.