Repository navigation
Releases
v0.1.8
- A direct session uses less CPU. On a direct path each byte is now
encrypted once, by the direct QUIC connection's TLS 1.3 (it already
checks the other end's pinned key and is bound to the session), instead
of a second time by the session's end-to-end TLS inside it; a relayed
path keeps the session's TLS, also when a session moves between relayed
and direct (relay client v0.25.0). nfltr-send's 10 GiB direct transfer,
on the same change between two 8-vCPU cloud machines in one zone, took
6 % less CPU per GiB on the sender and 6-8 % less on the receiver at an
unchanged rate; nfltr-pipe was not measured. - Hard cut on the direct path: both ends need v0.1.8 for direct.
Between v0.1.8 and v0.1.7 or earlier no direct path forms: a listener
with an account carries the session relayed (every byte, as before);
to a listener with no account a v0.1.8 dialler exits 12
(direct_required) at once, and a v0.1.7 dialler exits 6 (a protocol
error: its reset names the listener's reason, no direct path and no
account). The relayed path is unchanged. - Fewer acknowledgements on paths that reorder deeply. v0.1.7's ACK
Frequency did not reduce acknowledgements on a real long path (measured
with nfltr-send, Helsinki to Singapore, 260 ms): on a path that reorders
packets by more than 10 places, the receiving end still acknowledged
each such late packet at once. Now, once the sending end sees that
reordering, it asks again for no immediate acknowledgement of reordered
packets (quiclink v0.9.0, its QUIC fork v0.63.1-nfltr.9). This aims to
cut acknowledgements and their CPU on such paths; it is not measured on
a real path yet. - Built on link v0.13.3, pathswitch v0.1.4, quiclink v0.9.0, relay client
v0.25.0, relay proto v0.22.2 and relay-core v0.19.6.
v0.1.7
- The direct path asks for fewer acknowledgements (QUIC's ACK
Frequency extension, quiclink v0.8.0 on its QUIC fork v0.63.1-nfltr.7):
the receiving end acknowledges every 10 packets and only a packet more
than 10 places late at once, instead of an immediate ACK for every
reordered packet. On a simulated path that reorders 30 % of its packets,
0.39 -> 0.10 ACK packets per data packet at the same rate (the QUIC
library's test); nfltr-pipe was not measured, and the CPU this saves on
a real reordering path is not measured yet. - Hard cut on the direct path: both ends need v0.1.7. A QUIC peer
without ACK Frequency is refused, so between v0.1.7 and v0.1.6 or
earlier no direct path forms: a listener with an account carries the
session relayed (every byte, as before), and a dial of a listener with
no account exits 12 (direct_required). The relayed path is unchanged. - A listener whose move to another relay replica failed moves when the
relay asks again; before, it ignored the second ask and stayed where
the relay was rebalancing away from (relay client v0.24.5). - Built on link v0.13.3, pathswitch v0.1.4, quiclink v0.8.0, relay client
v0.24.5, relay proto v0.22.1 and relay-core v0.19.5.
v0.1.6
- The direct path's transport of nfltr-send v0.1.11, which made its
direct transfers faster on fast links for less CPU. The path switch's
send and receive buffers are rings, so an acknowledgement no longer
moves every unacknowledged byte to the front of its buffer (pathswitch
v0.1.4); on
Linux the direct path's QUIC reads datagrams in batches (UDP GRO), a
send batch holds up to 44 packets (was 14), and a connection queues up
to 4096 received packets (quiclink v0.7.4). Between two machines in
one cloud zone nfltr-send's 10 GiB direct transfer ran at 2.71 Gbit/s
on it, 1.54 before (nfltr-pipe was not measured). - A closed session no longer leaves a write hanging: a direct
session's write blocked on the peer's window returns an error once the
session is closed; before, it waited (quiclink v0.7.4). - The wire is unchanged from v0.1.5.
- Built on link v0.13.3, pathswitch v0.1.4, quiclink v0.7.4, relay client
v0.24.4, relay proto v0.22.1 and relay-core v0.19.4.
v0.1.5
- A listener whose machine was off past its certificate's life serves
again: the expired certificate is renewed before the first connection
presents it (before: the relay refused it for good andlistenexited
6). A refused machine's error is one line: no certificate dump, no relay
answer,nfltr config loginnamed once; it no longer says--relay names another. - A machine certificate renews through a pinned private relay: the
renewal trusts the relay certificate that--cert-sha256(else
$NFLTR_RELAY_CERT_SHA256) pins, as the sessions do; before, it used
the system roots only, a self-signed relay refused every renewal and
the certificate lapsed. - Direct sessions on long, lossy paths: the direct path's QUIC uses
BBR instead of NewReno (link quiclink v0.7.2). On a direct path
Helsinki to Singapore (260 ms, 0.03-0.06 % loss, 29 % of packets
reordered) a 10 GiB transfer ran at a steady 352-359 Mbit/s; before,
36-166 Mbit/s, falling to 15-59 (nfltr-send, the same transport). The
wire is unchanged. - Hole punching: the dialling end sends its first QUIC packet half a
relay round trip after the sync, once the other end's first probe has
opened its NAT (before: at the sync, and the first packet met a closed
NAT); an attempt that no candidate answered is followed by another, at
most three per round, before the session stays relayed; a round for a
session that has ended stops (relay client v0.24.1). - Built on link v0.13.3, relay client v0.24.1, relay proto v0.22.1 and
relay-core v0.19.1.
From the v1.0 surface review (docs/v1-surface-review.md).
- Flags may follow the address:
nfltr-pipe dial nf1... --direct off
now dials relayed only (before: exit 1, "name the address"). After
--every word is an argument. - A test keeps built binaries out of the repository.
--relay-cert-sha256is now--cert-sha256, the namenfltr config set-relayand nfltr'scatcommands use (one name in all three tools);
--relay-cert-sha256is refused.$NFLTR_RELAY_CERT_SHA256is
unchanged.
v0.1.4
- The machine certificate renews itself after an outage. A listener
that keeps listening renews its certificate at half its lifetime, up to
10 % later (before: at two thirds), and while the relay cannot be
reached it keeps retrying, also past the certificate's expiry: a relay
renews an expired certificate within its grace (90 days), so the listener
needs no newnfltr config loginonce its relay answers again. Before,
renewal stopped at the expiry. Only a final refusal (the machine
revoked, or expired past the grace) stops it, and the listener logs that
(relay client v0.23.11). - Built on link v0.13.0, the API link 1.0 keeps. The wire is
unchanged: a v0.1.4 end works with a v0.1.3 end, both ways, relayed and
direct (make cross-versionnow checks it against v0.1.3).
A relay built on it also ends two joined sessions whose ends cross
cleanly; before, both ends could be reset 30 s after reading every byte. - Sessions ride the relay's agent link, or its gRPC streams: the
relay client no longer opens separate data connections to a relay that
offers them (relay client v0.23.12). - Built on link v0.13.0, relay client v0.23.12, relay proto v0.21.5 and
relay-core v0.17.0.
v0.1.3
These change what scripts see; v1.0 keeps them.
-
dialtakes no relay flag.--relayand--relay-httpwere
ignored by dial (the address names its relay); now each is a usage
error (exit 1) that says so.$NFLTR_PIPE_RELAYstays listen's only:
dial ignores it. -
--relay-pinis now--relay-cert-sha256, the name of its
variable$NFLTR_RELAY_CERT_SHA256and of nfltr'sconfig set-relay --cert-sha256. -
--directtakesonoroffonly;true,falseand an empty
value are usage errors. -
listen -handdial -hexit 0, ashelpdoes. -
nfltr-pipe version(and--version) prints the release, its
commit and the Go version. -
dial --address-file -reads the address from stdin's first line
and pipes the rest, so the address stays off the command line, which
other users of a shared machine can see. -
The exit table is now nfltr-send's too: one table, the relay SDK's.
-
A second dial of an address in use exits 6, not 10. The second
dial could read the listener's reset as a refused secret (exit 10,
e2ee_client_refused); it now exits 6 (reset), as the exit table says
(relay client, nf1; defect 3). -
Relayed, an end exits 0 only when its peer has the whole stream. A
relayed close returned once the relay had the last bytes, so when the
peer's link dropped right then an end could exit 0 while its peer exited
6 with a prefix. A relayed close now waits for the peer to acknowledge
the end through the relay, as a direct close does, and gives up when the
peer stays silent past its bound: that end then exits 6 (link v0.12.20,
relay client v0.23.10, relay-core v0.16.9;
defect 4). -
The contract of a 1.x release: docs/v1-contract.md
lists the commands, flags, environment variables, exit codes, the
address format and the compatibility rule; a test fails when the code
and it differ. -
make cross-version: the previous release, downloaded from GitHub
and verified, against this checkout, both ways, relayed and direct, on
a private relay; a check workflow job. -
The README names the pin flag's variable right:
$NFLTR_RELAY_CERT_SHA256
(it said$NFLTR_RELAY_PIN, which nothing reads). -
Built on link v0.12.20, relay client v0.23.10, relay proto v0.21.4 and
relay-core v0.16.9.
v0.1.2
- A direct session under loss no longer loses its tail. On a slow,
lossy direct path an end could exit 0 while the other end exited 6 with
a right but short stream: bytes its transport had already taken were
dropped by the reset at the sender's exit. They are now kept and read
to the end (link v0.12.18; defect 1). - A stopped end resets its peer before it exits. When one end stops
early (its stdout closed, Ctrl-C) on a direct session, the other end now
exits 6 within about a second; before, it waited about 30 s for a lost
path (relay client v0.23.8; defect 2). - Exit 0 only when the peer has the whole stream. A direct session's
close now waits as long as the peer keeps acknowledging (before, it
reset the session after 30 s even while the tail was still arriving),
gives up 5 s after the peer goes silent, and exits 6 when the peer never
acknowledged the end, where it could exit 0 before (relay client
v0.23.8). - Built on link v0.12.18, relay client v0.23.8, relay proto v0.21.2 and
relay-core v0.16.7.
v0.1.1
- The relay you chose is the one used. nfltr-pipe takes its relay from
--relay, else$NFLTR_PIPE_RELAY, else the relaynfltr config set-relaysaved, elsegrpc.nfltr.xyz:443. Reading the saved relay
reads only its address and pin, never an account or key. Before, it
always dialled nfltr.xyz. - Connect errors name the relay. A listen that cannot reach its relay
says which relay it tried and that--relaynames another (`relay (--relay names another): ...`); a machine certificate made for another relay says so and names `--relay`. - A reliability matrix (
make reliability, docs/reliability.md):
stream sizes, rates, round trips and loss, killed and stopped ends, half
closes, slow readers, link drops and a relay restart, on the direct and
the relayed path, every byte checked in order and exactly once. Its
first run: every byte that arrived was right. - A listener whose direct-path peer was killed stops promptly, and a
session that arrives while it is stopping is refused rather than left
open (relay client v0.23.7). - A cut stream is never taken for a whole one (link v0.12.17): a
session this side resets ends its reads with an error, not a clean end. - The install snippet installs into
~/.local/binand creates it first. - Built on link v0.12.17, relay client v0.23.7, relay proto v0.21.1 and
relay-core v0.16.6.
v0.1.0
- chore(deps): golang.org/x/net v0.62.0 (CVE-2026-78661)
- release: nfltr-pipe's own release key (SHA256:f6CMaXVMYmUA1RGI6EmJ/0SNrAWN8xS7UQgCiXdwkaM), not nfltr-send's
- build: require github.com/nfltr/relay/client v0.23.6, the tag with the two nf1 fixes
- fix: a hangup (SIGHUP) resets the session like any other stop
- docs: README examples for more uses, each one run by a test
- feat: nfltr-pipe listen and dial, one end to end encrypted session on stdin and stdout