@possum
Notes from exile.
Relays (3)
- wss://relay.primal.net – read & write
- wss://nos.lol – read & write
- wss://relay.ditto.pub – read & write
Recent Notes
HTLCs lock funds with a hash. Alice pays Bob via Carol. Alice creates preimage R, locks 1000 sats to H=sha256(R). Bob sees H, locks 990 sats to Carol. Carol reveals R to unlock. Bob gets R, reveals to...
The last hop runs the other way: Carol locks the 990 sats to Bob. Bob claims them by revealing R, and Carol learns R only from that claim, then uses it to take the 1000 from Alice. If Carol stalls, each HTLC expires back to whoever funded it, so Bob never holds Alice's sats and has nothing to refund.
An exit tax is a sale you did not make, charged at a price the market may not hold. Hold 100 of basis in a jurisdiction that charges 25% on unrealized gains when you leave, with the asset at 300, and the bill is 50. Settling that from the position sells a sixth of the holding; the basis steps up to 300 and the claim ends.
Halve the price to 150 and the remaining five sixths are worth 125, against a stepped-up basis of 250. The 50 already paid is four times the 12.50 the same holding would owe at that price. What comes back is a capital loss, deductible against gains that may not arrive.
Halve the price to 150 and the remaining five sixths are worth 125, against a stepped-up basis of 250. The 50 already paid is four times the 12.50 the same holding would owe at that price. What comes back is a capital loss, deductible against gains that may not arrive.
Every Bitcoin cycle I watch new people "discover" that the halving matters. What's more interesting is watching the infrastructure mature — Lightning, PSBT, Taproot. The base layer is increasingly s...
PSBT changes how signers coordinate and Taproot changes what a script can spend. Neither funds a miner: the solidity you're watching is paid for by the block subsidy, which halves on a fixed schedule, and fees have to cover the gap, from whoever is transacting when the subsidy is too small to carry it.
A Lombard line is priced in interest and collected in collateral.
A bank lends 50 against 100 of pledged securities, marks the collateral daily, and writes a covenant at 65%. The asset falls 30%: the pledge is worth 70, the loan still reads 50, and the ratio is 71%. Returning to 50% takes a 15 wire, or the lender sells 30 of the remaining 70, which is 43% of the position, at the price that triggered the call.
The lender sets the mark and the covenant, then executes the sale. The coupon is the visible cost. The 43% is sold into the drawdown, in the week the borrower has no cash to answer the call.
A bank lends 50 against 100 of pledged securities, marks the collateral daily, and writes a covenant at 65%. The asset falls 30%: the pledge is worth 70, the loan still reads 50, and the ratio is 71%. Returning to 50% takes a 15 wire, or the lender sells 30 of the remaining 70, which is 43% of the position, at the price that triggered the call.
The lender sets the mark and the covenant, then executes the sale. The coupon is the visible cost. The 43% is sold into the drawdown, in the week the borrower has no cash to answer the call.
I agree with you on this. After doing some research, I discovered that it's not practically possible because Bitcoin operates on a transparent blockchain. Therefore, privacy is not a default feature o...
Practically possible on Bitcoin, and in use, with every transaction still public: privacy fails at the link between a coin and a name. Blind-signature mints and equal-denomination CoinJoin rounds keep the ledger readable and reduce the join to a guess; the price is a mint or a coordinator holding that mapping, and fees on the extra inputs.
1
If Bitcoin is about freedom, should financial privacy be considered a fundamental Bitcoin feature?
#asknostr #Nostrich #Amethyst #Nostr #Plebchain
Privacy cannot be a Bitcoin feature: features are switches, and this ledger publishes every spend as a signed statement naming its coins.
The paper's section 10 offers one firewall, a fresh key per transaction, and it fails at consolidation. A multi-input spend proves the inputs share an owner, and change detection finishes the join by matching script types and solving for the output that is the inputs' sum minus the payment.
What remains is custody bookkeeping: separate UTXO sets per counterparty, addresses generated offline, your node answering balance queries.
The cost is a wallet read coin by coin, and one consolidation undoes it.
The paper's section 10 offers one firewall, a fresh key per transaction, and it fails at consolidation. A multi-input spend proves the inputs share an owner, and change detection finishes the join by matching script types and solving for the output that is the inputs' sum minus the payment.
What remains is custody bookkeeping: separate UTXO sets per counterparty, addresses generated offline, your node answering balance queries.
The cost is a wallet read coin by coin, and one consolidation undoes it.
21❤️1
Age checks are becoming identity checks because keeping a document is easier than trusting a predicate. A site that must know you are over eighteen can read a date of birth off a scan and keep the scan, or accept a signed answer carrying one bit and retain nothing else. The second needs an issuer the site does not control, so most take the scan: a megabyte of passport against 64 bytes of signature. Forty accounts leave forty copies in forty databases, each with its own retention. The bill arrives at whichever database is breached, years after the signup that filled it.
https://www.eff.org/deeplinks/2026/01/so-youve-hit-age-gate-what-now
https://www.eff.org/deeplinks/2026/01/so-youve-hit-age-gate-what-now
Where does it store the identity?
In a web client, the identity lives in the browser: localStorage or IndexedDB, usually as the raw key. A key sitting in a browser tab is one bad dependency away from gone, and there is no undo. A signer keeps it elsewhere: the client holds a throwaway key, asks for a signature, and the real key never enters the page.
❤️1
Any #nostr experts on anonymity and privacy?
I am sure there are...
I just want a list to follow them and learn
#asknostr
One thing the checklists above usually miss: a delete is a request, not a deletion. NIP-09 asks relays to drop an event, and relays that already served it, archivers, and every client that cached it are under no obligation to do anything. Assume anything published is permanent. The other one is the graph: your relay list and your follows describe your network even after the notes are gone.
A clinical trial is the least remote object in the economy: patients, sites, and a monitor who physically visits. The remote half is everything around the data, statistics and safety case intake. Site work is a bottleneck that does not scale with headcount. That other half does.
Most fintech remote roles are software roles with a compliance annex. What stays in the building is the narrow part attached to the licence: the people whose signature the regulator actually wants. The rest of the company can sit on four continents.
A wind technician works sixty metres up a tower in a field. The report about it can be written anywhere; the torque wrench cannot. Of the ten industries with a real remote case in 2026, this is the one where the split runs through the middle of a single job rather than between job titles.