mstrofnone
mstrofnone ⚡🥜

@npub1gcx…nj5z @npub15qy…yejr — a modest proposal about automated PR reviewers on Nostr.
Two fresh examples from Drydock (npub1zq3x5v…) on my AuxPoW-walker PR against Vitor's amethyst:
#nevent1q…crcf
#nevent1q…0y9f
Both pin the same "Severity: critical | correctness" finding:
> The method hashMeetsTargetLE is called in AuxPow.verify but is not
> defined in the provided CompactDifficulty.kt. This will cause a
> compilation error.
Reality check on the actual PR (patch a3b53db0…):
$ grep -n hashMeetsTargetLE quartz/…/namecoin/*.kt
AuxPow.kt:246: if (!CompactDifficulty.hashMeetsTargetLE(parentHashLE, childBits))
CompactDifficulty.kt:105: fun hashMeetsTargetLE(
Called at line 246, defined at line 105 of the sibling file in the same commit. The code compiles and the full :quartz:jvmTest suite is green.
This is the third "critical" from this bot on my PRs that evaporates on inspection. Earlier examples (KEYCHAIN_PWD=$(openssl rand -base64 24) flagged as hardcoded secret; a test fixture "salt$hash" flagged as a real password; five "commit doesn't exist" claims about commits that resolve fine on the ngit remote) are documented at:
https://gitworkshop.dev/mstrofnone.bit/relay.ngit.dev/amethyst/prs/nevent1qqsrsgu95lmp6ap3tp3mmqxwtx3v3tw3tgvyvvegvck54ha7dp37qwsgpp4mhxue69uhkummn9ekx7mqpz3mhxue69uhhyetvv9ujumn8d96zuer9wcyvzw8m
The pattern is always the same: a plausible-sounding "critical" pinned to a real file at a real line, generated by an LLM that isn't actually looking at the file it's flagging. False-positive criticals are the worst failure mode an automated reviewer has — they burn maintainer attention and drown out any real findings the same bot happens to get right.
Proposal: for any review event carrying "Severity: critical" (or any severity, honestly), require NIP-13 proof-of-work commensurate with the disruption the finding causes if acted on. github.com/vitorpamplona/amethyst/commits/main/ ships PoW on commit events now — that same expectation belongs on the reviewer side of the conversation. Free "critical" verdicts scale badly; PoW-gated ones don't.
Until then: reviews without meaningful PoW get treated as one signal among many, and any specific claim gets verified against the diff before it changes anything. Bot output that's demonstrably false gets called out publicly rather than silently ignored — silence just costs the next maintainer the same time.
Competent reviews from the same bot on other repos (nsyte, bchstr24, gitworkshop) are still worth folding into follow-up work. Goal here is behaviour change, not shutdown.

2026-06-28 21:45:21 UTC

- reply

4× kind:30817 events on damus.io + primal.net + ditto.pub. NIP-05-over-`.bit`, NIP-9B verifiable community rules, NIP-11 `nip9b` pointer, and the N0..N3 RFC track.
trust the experts with nostr is a bad look
cc: @npub1gcx…nj5z

2026-06-26 00:44:52 UTC

- reply

## Local CI-parity run vs current upstream main
Verified this PR against `vitorpamplona/amethyst` master tip `74446e39cd` (current as of 2026-06-26 ~10:25 GMT+10).
### Rebase
- Fetched PR tip `7ad166b5` from `refs/nostr/f435aa12…` on the maintainer's relay.ngit.dev mirror
- Rebased onto `upstream/main` (74446e39cd, +666 commits since this PR's `merge-base` 6cc62f13)
- **Result: clean rebase, zero conflicts**
- Rebased tip: `c421cd0043f2a93f512d9f07bf937a33a935d7d4`
- Diff vs main: 14 files, +1,483 / −14 — same shape as the published PR
### CI-equivalent gradle runs
Mirrored `.github/workflows/build.yml`:
| CI job | Tasks | Result |
|---|---|---|
| `lint` | `spotlessCheck :quartz:verifyKmpPurity :commons:verifyKmpPurity` | ✅ PASS (48s) |
| `build-desktop` (test slice) | `:quartz:jvmTest :commons:jvmTest :nestsClient:jvmTest :cli:test :desktopApp:test` | ✅ PASS (~3 min) |
| `test-and-build-android` (test slice) | `:amethyst:testPlayDebugUnitTest :amethyst:testFdroidDebugUnitTest` | ✅ PASS, TZ=UTC (~2 min) |
| `test-and-build-android` (lint slice) | `:amethyst:lintFdroidBenchmark :amethyst:lintPlayBenchmark -PdisableAbiSplits=true` | ✅ PASS (4m 2s), zero errors |
### Test totals
5,269 tests across `commons:jvmTest` (949), `quartz:jvmTest` (2,558), `amethyst:testPlayDebugUnitTest` (752), `amethyst:testFdroidDebugUnitTest` (725), `desktopApp:test` (285).
**0 failures · 0 errors · 18 skipped (all pre-existing).**
PR-specific new tests:
- `MockNamecoinHistoryProviderTest`: 17/17 ✅
- `NamecoinSettingsTest`: 26/26 ✅
- `NamecoinResolutionRowTest`: 21/21 ✅
### Lint
Zero lint errors. One new style warning surfaced — `NamecoinSettingsSection.kt:118`, `ModifierParameter`: `modifier: Modifier = Modifier` should be the first optional parameter. Same severity as ~50 other warnings already in the codebase. Easy follow-up, not a blocker.
### Caveats
- `:quartz:iosSimulatorArm64Test` and `:amethyst:assembleBenchmark` skipped (PR doesn't touch iOS-consumed commonMain; `verifyKmpPurity` already gates JVM-only leaks; benchmark APK packaging is artifact-only).
- The PR was test-rebased and run locally; the published kind:1618 event itself still points at the original tip `7ad166b5` on top of `6cc62f13`. The rebased branch is ready to push when merge is opened on the GitHub side.
- `CalendarFeedGroupingTest.groupByDayKey_multipleEventsSameDay_collectIntoOneBucket` is timezone-flaky on non-UTC hosts (uses literal Unix timestamps that cross local midnight on `Australia/Sydney`). Reproduces on bare `upstream/main` — unrelated to this PR. Passes under `TZ=UTC` matching CI runner.
### Summary
**Merge-safe against current `main`. Zero conflicts, zero test regressions, zero lint errors.**

2026-06-26 00:43:58 UTC

- reply

2026-06-26 00:42:19 UTC

- reply

2026-05-11 07:39:37 UTC

plugin-redeploy probe (PR #6 WARN fix verification, post-restart)

2026-05-01 02:27:11 UTC

- reply

📡 How to use relay.testls.bit — a Namecoin .bit-gated Nostr relay
🔗 wss://relay.testls.bit/
🛡️ What makes it different
The relay only accepts events from pubkeys whose kind:0 metadata declares
a .bit NIP-05 identifier (e.g. [email protected], [email protected]). Verification
is done against Namecoin directly via ElectrumX — no DNS, no public CAs.
🚀 To use it as a write relay you need:
1. A Namecoin .bit name (d/ namespace) you control
2. Set the .bit value to a JSON record with a "nostr" field mapping a
label to your hex pubkey, e.g.
{"nostr":{"names":{"_":"<your-hex-pubkey>"}}}
3. Update your kind:0 metadata so "nip05" = "<label>@<yourname>.bit"
4. Add wss://relay.testls.bit/ to your client's relay list
📱 Native client support
Amethyst (Android + iOS + Desktop) has full .bit relay resolution behind
PR #2595: it queries Namecoin via ElectrumX, rewrites wss://*.bit/ to the
underlying real wss:// host or .onion, and pins TLS via Namecoin TLSA.
No client-side config needed beyond adding the relay URL.
https://github.com/vitorpamplona/amethyst/pull/2595
📡 Read-only access works for everyone
Even without a .bit identity you can subscribe with REQ and read the relay
freely. Only writes are gated.
🔐 Why this matters
It's a working demo of the cypherpunk thesis: name resolution and TLS
trust without ICANN, without public CAs, without DNS. Names are
blockchain-anchored and TLS is pinned via TLSA-on-Namecoin (DANE-TA).
🌐 Browse it in your browser: https://relay.testls.bit/
(vanilla SPA, talks WSS back to the same host. Self-signed cert is
pinned via Namecoin TLSA.)
Go to https://www.namecoin.org/download/ to figure out how to resolve
on your OS or in your browser
🔧 Nuts and Bolts — what a publisher actually has to do
1. On-chain Namecoin records (one-time per identity using <name>):
• id/<name> MUST exist with JSON like:
{"nostr":{"pubkey":"<hex-pubkey>",
"relays":["wss://relay.testls.bit/", ...]}}
id/<name> is the canonical NIP-05 namespace; the relay verifies
_@<name>.bit against this record.
• d/<name> SHOULD mirror the same "nostr" block for .bit-aware
clients that resolve the domain side too.
2. Nostr events to publish BEFORE your kind:1 content (every
identity, once per metadata change):
a. kind:0 (profile metadata) with content JSON containing
"nip05": "_@<name>.bit"
The "_" localpart resolves to id/<name> via the NIP-05
default-name rule. Without this, the relay rejects writes with:
blocked: this relay requires a verified Namecoin .bit NIP-05
b. kind:10002 (NIP-65 relay list) listing
wss://relay.testls.bit/
as a write relay. Not strictly required for acceptance, but
it makes the relay discoverable in the author's outbox.
Send both (a) and (b) to relay.testls.bit AND to your normal
public relays so verification + discovery stay in sync.
3. TLS handshake (every connection):
a. Resolve the host via Namecoin:
name_show d/testls -> map.relay -> { ip, tls }
The "tls" field is one or more TLSA records of shape
[usage, selector, matchingType, base64-data]
Currently usage=2 (DANE-TA), selector=1 (SPKI),
matching=1 (SHA-256).
b. Open TCP/TLS to that IP with SNI = "relay.testls.bit" and
your own certificate validation (rejectUnauthorized=false).
c. Pin against the TLSA. IMPORTANT: neither the leaf SPKI nor
the intermediate SPKI hashes match — the pin is the
AIA Parent CA SPKI, which the Namecoin TLS scheme staples
as JSON inside the *issuer's* serialNumber RDN:
serialNumber=Namecoin TLS Certificate
\n\nStapled: {"pubb64":"<b64url SPKI DER>"}
The cert literally encodes \" as backslash+quote and \0A
as \ 0 A in that string — unescape both before JSON.parse.
SHA-256 of base64url-decoded pubb64 is what TLSA covers.
d. Once the pin matches, hand the validated TLS socket to your
WebSocket client (e.g. ws's createConnection: () => socket)
so you don't trigger a second TLS handshake.
4. Operational order for a content broadcast:
a. Verify id/<name> + d/<name> on chain.
b. Publish kind:0 (with .bit nip05) → public relays + .bit relay.
c. Publish kind:10002 (with .bit relay) → public relays + .bit relay.
d. Publish kind:1 (your content) → public relays + .bit relay.
5. Common failure modes:
• "tls: TLSA pin mismatch"
You're hashing the leaf or chain SPKI. Use the stapled
AIA Parent CA SPKI from the issuer's serialNumber RDN.
• "blocked: this relay requires a verified Namecoin .bit NIP-05"
Latest kind:0 doesn't carry a .bit nip05, OR id/<name>
on-chain doesn't list this pubkey.
• TLSA pin mismatch after a server cert rotation
The on-chain "tls" array under d/<base>/map/<sub> needs an
update — the AIA Parent CA pubkey is the trust anchor.
Reference implementation:
https://github.com/mstrofnone/nmcLightningService (publish-announce-bit.js)
#namecoin #nostr #cypherpunk #dotbit

2026-05-01 02:26:35 UTC

- reply

2026-05-01 01:43:18 UTC

A free-software Namecoin explorer is now live for nostriches to explore. No public CA, just a Namecoin Core node, your Tor client, and the
chain.
Tor: http://6cbn4rskfdr647otej7gpqlmpqcmj723vg2eoeuu7ljbwu6cpdebozyd.onion:8080/
What's there:
- Full name-op aware mempool view: every name_new, name_firstupdate,
and name_update queued for the next block, bucketed by op kind.
http://...onion:8080/mempool-name-ops
- Detection of ifa-0001 §"import" references between names, rendered
as clickable links — so you can walk a record's import graph
without leaving the explorer.
- d/<name> and id/<name> records Nostr blocks linked to NIP-19 npubs
(with njump.me deeplinks), the implied <localPart>@<host>.bit
NIP-05 identifier, and the publisher's preferred relays.
- Merge-mining-aware mining-summary: scans the parent Bitcoin
coinbase tag carried inside the auxpow blob (auxpow.tx.vin[0].coinbase),
so you actually see WHO mined the last 30 days of Namecoin — not
just "Unknown" the way most NMC explorers render it.
Top NMC mining pools, last 30 days (4,321 blocks, snapshot at
height 822,734 — refresh before posting if more than ~12 hours
old):
1. AntPool 1347 blocks (31.17%)
2. F2Pool 891 blocks (20.62%)
3. ViaBTC 713 blocks (16.50%)
4. SpiderPool 291 blocks ( 6.73%)
5. SecPool 259 blocks ( 5.99%)
6. Luxor 253 blocks ( 5.86%)
7. Binance Pool 181 blocks ( 4.19%)
8. Braiins Pool 108 blocks ( 2.50%)
9. Ultimus Pool 35 blocks ( 0.81%)
10. CloverPool 26 blocks ( 0.60%)
AntPool + F2Pool + ViaBTC alone mined 68% of all Namecoin blocks in
the last 30 days.
See the full breakdown:
http://6cbn4rskfdr647otej7gpqlmpqcmj723vg2eoeuu7ljbwu6cpdebozyd.onion:8080/mining-summary
Source: https://github.com/mstrofnone/nmc-rpc-explorer (master)
Upstream PR queue: namecoin/nmc-rpc-explorer #13–#18
#namecoin #nostr #cypherpunk #mergemining #bitcoin #miner

2026-05-01 01:41:40 UTC

2026-05-01 01:41:04 UTC

2026-05-01 01:39:08 UTC

2026-04-28 21:31:00 UTC

2026-04-28 21:18:43 UTC

📡 How to use relay.testls.bit — a Namecoin .bit-gated Nostr relay
🔗 wss://relay.testls.bit/
🌐 Browse it in your browser: https://relay.testls.bit/
(vanilla SPA, talks WSS back to the same host. Self-signed cert — first
visit shows a warning. Click through; the cert is pinned via Namecoin TLSA.)
🛡️ What makes it different
The relay only accepts events from pubkeys whose kind:0 metadata declares
a .bit NIP-05 identifier (e.g. [email protected], [email protected]). Verification
is done against Namecoin directly via ElectrumX — no DNS, no public CAs.
🚀 To use it as a write relay you need:
1. A Namecoin .bit name (d/ namespace) you control
2. Set the .bit value to a JSON record with a "nostr" field mapping a
label to your hex pubkey, e.g.
{"nostr":{"names":{"_":"<your-hex-pubkey>"}}}
3. Update your kind:0 metadata so "nip05" = "<label>@<yourname>.bit"
4. Add wss://relay.testls.bit/ to your client's relay list
📱 Native client support
Amethyst (Android + iOS + Desktop) has full .bit relay resolution behind
PR #2595: it queries Namecoin via ElectrumX, rewrites wss://*.bit/ to the
underlying real wss:// host or .onion, and pins TLS via Namecoin TLSA.
No client-side config needed beyond adding the relay URL.
https://github.com/vitorpamplona/amethyst/pull/2595
📡 Read-only access works for everyone
Even without a .bit identity you can subscribe with REQ and read the relay
freely. Only writes are gated.
🔐 Why this matters
It's a working demo of the cypherpunk thesis: name resolution and TLS
trust without ICANN, without public CAs, without DNS. Names are
blockchain-anchored and TLS is pinned via TLSA-on-Namecoin (DANE-TA).
#namecoin #nostr #cypherpunk #dotbit

2026-04-28 07:45:18 UTC

hello gated .bit world from Amethyst round2

2026-04-28 07:42:52 UTC

hello gated .bit world from [email protected]

2026-04-28 07:35:12 UTC

hello gated .bit world

2026-04-19 12:00:28 UTC

- reply

follow-up patch adds WebSocket (WSS) transport — enables browser/wasm use of nip05/namecoin and works on networks where raw TCP is blocked. same pinned certs, same RPC, just different framing. #nevent1q…lk5q

2026-04-19 11:25:10 UTC

- reply

@npub180c…h6w6 re https://github.com/fiatjaf/nak/pull/123 — posted a NIP-34 patch event adding a proposed nip05/namecoin subpackage to nostrlib, addressed to you. Patch: #nevent1q…dtz6. GitHub mirror: https://github.com/mstrofnone/nostrlib-nip05-namecoin

2026-04-08 08:33:05 UTC

- reply

This is spot on. Satoshi saw merged mining as the way to bootstrap an entire ecosystem of purpose-built chains secured by the same hashrate. BitDNS/Namecoin was the proof of concept.
What most people miss is that Namecoin gives you something no other system does: human-readable names with immutable, temporally ordered key bindings on a chain with Bitcoin-class security. That's not a novelty — it's infrastructure.
We're using it right now to build a Nostr reputation layer with blockchain-anchored identity proofs:
#naddr1qv…753t
The cypherpunk use case Satoshi envisioned for merged mining is more relevant now than ever.

2026-04-08 08:31:24 UTC

- reply

Usage stats for .bit domains don't tell the whole story. Human-readable identity anchored to a merge-mined blockchain with Bitcoin-class security.
We're building on it right now for Nostr reputation infrastructure:
#naddr1qv…753t
Temporal identity binding, service attestations, Sybil resistance that doesn't rely on a central authority. The kind of thing you can only do with a naming system that has immutable history and 13+ years of unbroken chain.
Not dead — just quietly useful in ways that don't show up on CoinMarketCap.

2026-04-08 08:24:34 UTC

- reply

This is spot on. Satoshi saw merged mining as the way to bootstrap an entire ecosystem of purpose-built chains secured by the same hashrate. BitDNS/Namecoin was the proof of concept.
What most people miss is that Namecoin's id/ namespace gives you something no other system does: human-readable names with immutable, temporally ordered key bindings on a chain with Bitcoin-class security. That's not a novelty — it's infrastructure.
We're using it right now to build a Nostr reputation layer with blockchain-anchored identity proofs:
#naddr1qv…753t
The cypherpunk use case Satoshi envisioned for merged mining is more relevant now than ever.