commits · hyperswarm-peer

hyperdht: dedup inbound handshakes by their Noise message, as hyperdht does

ed8df3a by plankatron · 2026-10-01

dht_rpc: read the query id BEFORE _advance_query

7177173 by plankatron · 2026-10-01

udx: a PeerConnection outlives its stream's libuv handles (uv__finish_close assert)

3c79be0 by plankatron · 2026-09-30

udx: an opt-in stream consumer — the secret-stream without our Protomux

8b811f5 by plankatron · 2026-09-30

udx: say when a peer is dropped for silence, and when keepalives start

272d72c by plankatron · 2026-09-29

swarmd-ctl: control a running daemon, instead of hunting for its process

50dc99a by plankatron · 2026-09-11

swarmd/build.sh: the toolchain is a variable, because a plain build is not portable

ec0c817 by plankatron · 2026-09-10

udx: liveness — an idle peer is now told about, and it is the reference's wire

9ae462a by plankatron · 2026-09-10

swarmd-session: session-daemon's socket, swarmd's insides

9de9ef8 by plankatron · 2026-09-10

dht: the relay stage gets JS's 15s deadline, not the 30s backstop

0e31891 by plankatron · 2026-09-10

docs: gate status after D5/D7, and relink what the transport change touched

38fe11b by plankatron · 2026-09-10

swarmd: gates D5 and D7 — one passes, one found a hole three layers deep

6fb11ed by plankatron · 2026-09-10

docs: step 5 is done — mark it, and show how to use it

e4b70e5 by plankatron · 2026-09-10

build: relink the gates that carry udx_transport

858537d by plankatron · 2026-09-10

swarmd-probe: rebuilt on the client library, and it got shorter

07e2b42 by plankatron · 2026-09-10

swarmd: a C++ client library, so the wire has ONE implementation

54eb4c5 by plankatron · 2026-09-10

swarmd: HELLO answered with two zeroes and a comment saying someone else would fill them

af54824 by plankatron · 2026-09-10

swarmd: content teardown freed the index out from under the loop

01e8400 by plankatron · 2026-09-10

udx: the loop-wake guard existed at ONE of twenty call sites

7a55292 by plankatron · 2026-09-10

swarmd: the new gate binary was tracked; every other one here is not

607b742 by plankatron · 2026-09-10

docs: what retiring session-daemon actually needs, and why not here

bda998f by plankatron · 2026-09-10

docs: price the rest of defect B — both halves, from the reference's constants

b0cf7a1 by plankatron · 2026-09-10

test: A2 is not reproducible — the measurement, and the code it sent back

fb3a85c by plankatron · 2026-09-10

dht: a holepunch reply that does not decode fails the attempt now, not in 30s

f6e430d by plankatron · 2026-09-10

swarmd: the content plane — 100 MB by hash, 996 bytes over the socket

62db3a6 by plankatron · 2026-09-10

liboverlay: cancel() and progress(), and the landmine the obvious cancel lays

e6a0fbd by plankatron · 2026-09-10

build: rebuild every committed binary against the libcompact codec

90d557d by plankatron · 2026-09-10

compact_encoding: forward to libcompact instead of writing the bytes again

cb5ff99 by plankatron · 2026-09-10

swarmd: the transport object cache compared against the wrong file

67d40c6 by plankatron · 2026-09-10

test: pin what a decoder owes its caller on a TRUNCATED frame

5f3f980 by plankatron · 2026-09-10

test: four gates the build script never built

d5c4bda by plankatron · 2026-09-10

swarmd: the two-daemon crossing, proven on the public DHT

f826718 by plankatron · 2026-09-10

swarmd-probe: offer the frame until it is taken, not once on a timer

5e13f1f by plankatron · 2026-09-10

swarmd: SEND answered about the queue, not about the frame

932528c by plankatron · 2026-09-10

swarmd: channels, and a C++ probe that speaks the wire with generated encoders

ce9311e by plankatron · 2026-09-10

swarmd: it runs — transport attached, and the UDX thread stays off the socket

401d7fb by plankatron · 2026-09-10

swarmd: the control plane, dispatched by generated code end to end

1818f6d by plankatron · 2026-09-10

swarmd: the local socket, with the buffering session-daemon does not have

b630b7f by plankatron · 2026-09-10

swarmd: generate the wire in C, and prove the toolchain compiles

e8c5c1d by plankatron · 2026-09-10

docs: write the swarmd design down before it is lost

8c1e7e2 by plankatron · 2026-09-10

test: verify the holder path against a real testnet, and assert the token

96bcaea by plankatron · 2026-09-10

dht: ping the nodes that hold our record, and know which ones they are

afd59cb by plankatron · 2026-09-10

docs: we do not know which nodes hold our record — B's ping needs this first

edc04b7 by plankatron · 2026-09-10

test: a JS control, and our connect-failure rate turns out to be the network's

829553c by plankatron · 2026-09-10

swarm: record a peer's topic before the already-connected early return

4b09bd3 by plankatron · 2026-09-10

dht: keep our own record alive — the self-announce happened once, ever

5974dc0 by plankatron · 2026-09-10

dht: fail a connect when find_peer exhausts, not 30 seconds later

261c01c by plankatron · 2026-09-10

README: the TODO list was contradicted by 27 bound methods

b906e0c by plankatron · 2026-09-10

docs: correct §11 — JS's connect release is 10-15s, not immediate

4c2e9d9 by plankatron · 2026-09-10

dht: the handshake pool fix, verified on the public DHT

c95ac63 by plankatron · 2026-09-10

hyperswarm-peer / commits