scaffold: M0 skeleton + agent onboarding (build, architecture, progress)
Turn the design into a buildable, dependency-free M0 skeleton plus the
onboarding layer so a new agent can pick up instantly.
Build system:
- CMake + CMakePresets (dev = no deps; server-release = vcpkg) + vcpkg.json
- Skeleton builds with just a C++20 compiler; deps stay off until needed
- .gitattributes (LF), .gitignore, .clang-format
Core (libvoicecat):
- core/include/voicecat.h: full C ABI (the client/server contract), stubbed
- core/proto/voicecat.proto: control-plane wire format, matches docs/protocol.md
- src/{net,crypto,codec,protocol,session,audio,core}: subsystem stubs that
return VC_ERR_NOT_IMPLEMENTED, each pointing to its design doc
- server/ (voicecat-server) and tools/vccli/ link the core
- tests/: CTest smoke test asserting the C ABI contract (behavior, not just build)
- clients/{apple,windows}: M4 placeholders
Onboarding for agents:
- CLAUDE.md: hub — build/test commands, architecture at a glance, doc map, rules
- AGENTS.md: working method (behavior-driven; clean compile is the floor not the goal)
- PROGRESS.md: living tracker — M0 done, M1 task checklist, "where we left off"
Verified: cmake --preset dev && cmake --build --preset dev && ctest --preset dev → green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-15 21:09:09 +02:00
|
|
|
#include "net/transport.h"
|
|
|
|
|
|
feat(M1): TCP/TLS control plane -- auth, channels, ephemeral text
Implements the full M1 milestone. Two clients authenticate over TLS 1.3
(guest + Argon2id password) and exchange channel + private text messages
through a real server. All five ctest --preset m1-dev tests pass in ~1 s.
Key components added:
- vcpkg baseline + m1-dev preset (protobuf/mbedTLS/libsodium/asio/sqlite3)
- FrameCodec feed+emit, encode/decode_envelope, protobuf codegen
- TcpServerConn with blocking TLS handshake thread + tls_read_loop
- TlsContext (mbedTLS 1.3, ECDSA-P256 self-signed cert, TOFU on client)
- WorkerPool (3 threads, used for Argon2id)
- Database: SQLite + libsodium Argon2id, account lifecycle, bootstrap admin
- ServerIdentityManager: Ed25519 key + cert generate/persist/fingerprint
- ConnSession state machine: WaitingHello -> WaitingAuth -> Authenticated
- SessionRegistry: channel tree, user map, text routing, broadcast
- vc_client full M1 C ABI: connect/TLS/handshake/auth/text/disconnect
- voicecat-admin CLI: account add/reset/del/list
- test_m1_integration: M1 exit criterion, verified green
Bug fixed: double-framing in ConnSession::send_envelope -- encode_envelope
was adding the [4-byte len] prefix, then TcpServerConn::send_frame added
a second one, causing the client to parse [len][proto] as protobuf (silent
failure). Fixed by serializing raw protobuf bytes in send_envelope and
letting send_frame apply the single length prefix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-15 23:48:44 +02:00
|
|
|
#ifdef VOICECAT_HAS_NET
|
|
|
|
|
|
|
|
|
|
#include <cstring>
|
|
|
|
|
|
|
|
|
|
#include "crypto/crypto.h"
|
|
|
|
|
|
scaffold: M0 skeleton + agent onboarding (build, architecture, progress)
Turn the design into a buildable, dependency-free M0 skeleton plus the
onboarding layer so a new agent can pick up instantly.
Build system:
- CMake + CMakePresets (dev = no deps; server-release = vcpkg) + vcpkg.json
- Skeleton builds with just a C++20 compiler; deps stay off until needed
- .gitattributes (LF), .gitignore, .clang-format
Core (libvoicecat):
- core/include/voicecat.h: full C ABI (the client/server contract), stubbed
- core/proto/voicecat.proto: control-plane wire format, matches docs/protocol.md
- src/{net,crypto,codec,protocol,session,audio,core}: subsystem stubs that
return VC_ERR_NOT_IMPLEMENTED, each pointing to its design doc
- server/ (voicecat-server) and tools/vccli/ link the core
- tests/: CTest smoke test asserting the C ABI contract (behavior, not just build)
- clients/{apple,windows}: M4 placeholders
Onboarding for agents:
- CLAUDE.md: hub — build/test commands, architecture at a glance, doc map, rules
- AGENTS.md: working method (behavior-driven; clean compile is the floor not the goal)
- PROGRESS.md: living tracker — M0 done, M1 task checklist, "where we left off"
Verified: cmake --preset dev && cmake --build --preset dev && ctest --preset dev → green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-15 21:09:09 +02:00
|
|
|
namespace voicecat::net {
|
|
|
|
|
|
feat(M1): TCP/TLS control plane -- auth, channels, ephemeral text
Implements the full M1 milestone. Two clients authenticate over TLS 1.3
(guest + Argon2id password) and exchange channel + private text messages
through a real server. All five ctest --preset m1-dev tests pass in ~1 s.
Key components added:
- vcpkg baseline + m1-dev preset (protobuf/mbedTLS/libsodium/asio/sqlite3)
- FrameCodec feed+emit, encode/decode_envelope, protobuf codegen
- TcpServerConn with blocking TLS handshake thread + tls_read_loop
- TlsContext (mbedTLS 1.3, ECDSA-P256 self-signed cert, TOFU on client)
- WorkerPool (3 threads, used for Argon2id)
- Database: SQLite + libsodium Argon2id, account lifecycle, bootstrap admin
- ServerIdentityManager: Ed25519 key + cert generate/persist/fingerprint
- ConnSession state machine: WaitingHello -> WaitingAuth -> Authenticated
- SessionRegistry: channel tree, user map, text routing, broadcast
- vc_client full M1 C ABI: connect/TLS/handshake/auth/text/disconnect
- voicecat-admin CLI: account add/reset/del/list
- test_m1_integration: M1 exit criterion, verified green
Bug fixed: double-framing in ConnSession::send_envelope -- encode_envelope
was adding the [4-byte len] prefix, then TcpServerConn::send_frame added
a second one, causing the client to parse [len][proto] as protobuf (silent
failure). Fixed by serializing raw protobuf bytes in send_envelope and
letting send_frame apply the single length prefix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-15 23:48:44 +02:00
|
|
|
// ── TcpControlChannel ────────────────────────────────────────────────────────
|
|
|
|
|
|
|
|
|
|
TcpControlChannel::TcpControlChannel(TcpChannelCallbacks cbs)
|
|
|
|
|
: work_guard_(asio::make_work_guard(io_)),
|
|
|
|
|
socket_(io_),
|
|
|
|
|
strand_(io_.get_executor()),
|
|
|
|
|
cbs_(std::move(cbs)) {
|
|
|
|
|
net_thread_ = std::thread([this] { run_loop(); });
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
TcpControlChannel::~TcpControlChannel() { close(); }
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::run_loop() { io_.run(); }
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::async_connect(const std::string& host, uint16_t port) {
|
|
|
|
|
auto resolver = std::make_shared<asio::ip::tcp::resolver>(io_);
|
|
|
|
|
resolver->async_resolve(
|
|
|
|
|
host, std::to_string(port),
|
|
|
|
|
[this, resolver](std::error_code ec, asio::ip::tcp::resolver::results_type eps) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
if (cbs_.on_connect_error) cbs_.on_connect_error(ec);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
asio::async_connect(socket_, eps,
|
|
|
|
|
[this](std::error_code ec2, const asio::ip::tcp::endpoint&) {
|
|
|
|
|
if (ec2) {
|
|
|
|
|
if (cbs_.on_connect_error) cbs_.on_connect_error(ec2);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
connected_.store(true, std::memory_order_release);
|
|
|
|
|
if (cbs_.on_connected) cbs_.on_connected();
|
|
|
|
|
start_read();
|
|
|
|
|
});
|
|
|
|
|
});
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::send_frame(std::vector<uint8_t> payload) {
|
|
|
|
|
std::vector<uint8_t> wire;
|
|
|
|
|
protocol::FrameCodec::emit(payload, wire);
|
|
|
|
|
asio::post(strand_, [this, w = std::move(wire)]() mutable {
|
|
|
|
|
send_queue_.push_back(std::move(w));
|
|
|
|
|
if (!sending_) do_send();
|
|
|
|
|
});
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::do_send() {
|
|
|
|
|
if (send_queue_.empty()) { sending_ = false; return; }
|
|
|
|
|
sending_ = true;
|
|
|
|
|
auto& front = send_queue_.front();
|
|
|
|
|
asio::async_write(socket_,
|
|
|
|
|
asio::buffer(front),
|
|
|
|
|
asio::bind_executor(strand_,
|
|
|
|
|
[this](std::error_code ec, std::size_t) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(ec);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
send_queue_.pop_front();
|
|
|
|
|
do_send();
|
|
|
|
|
}));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::start_read() {
|
|
|
|
|
asio::async_read(socket_, asio::buffer(len_buf_, 4),
|
|
|
|
|
[this](std::error_code ec, std::size_t n) { handle_length(ec, n); });
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::handle_length(std::error_code ec, std::size_t) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
if (ec == asio::error::eof || ec == asio::error::connection_reset) {
|
|
|
|
|
if (cbs_.on_disconnected) cbs_.on_disconnected();
|
|
|
|
|
} else {
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(ec);
|
|
|
|
|
}
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
uint32_t length =
|
|
|
|
|
(static_cast<uint32_t>(len_buf_[0]) << 24) |
|
|
|
|
|
(static_cast<uint32_t>(len_buf_[1]) << 16) |
|
|
|
|
|
(static_cast<uint32_t>(len_buf_[2]) << 8) |
|
|
|
|
|
static_cast<uint32_t>(len_buf_[3]);
|
|
|
|
|
|
|
|
|
|
if (length > protocol::kMaxFrameBytes) {
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(asio::error::message_size);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
if (length == 0) {
|
|
|
|
|
if (cbs_.on_frame) cbs_.on_frame({});
|
|
|
|
|
start_read();
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
body_buf_.resize(length);
|
|
|
|
|
asio::async_read(socket_, asio::buffer(body_buf_),
|
|
|
|
|
[this, length](std::error_code ec, std::size_t n) { handle_body(length, ec, n); });
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::handle_body(uint32_t, std::error_code ec, std::size_t) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
if (ec == asio::error::eof || ec == asio::error::connection_reset) {
|
|
|
|
|
if (cbs_.on_disconnected) cbs_.on_disconnected();
|
|
|
|
|
} else {
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(ec);
|
|
|
|
|
}
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
if (cbs_.on_frame) cbs_.on_frame(body_buf_);
|
|
|
|
|
start_read();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpControlChannel::close() {
|
|
|
|
|
if (closing_.exchange(true)) return;
|
|
|
|
|
asio::post(io_, [this] {
|
|
|
|
|
std::error_code ignored;
|
|
|
|
|
socket_.shutdown(asio::ip::tcp::socket::shutdown_both, ignored);
|
|
|
|
|
socket_.close(ignored);
|
|
|
|
|
});
|
|
|
|
|
work_guard_.reset();
|
|
|
|
|
if (net_thread_.joinable()) net_thread_.join();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// ── TcpServerConn ────────────────────────────────────────────────────────────
|
|
|
|
|
|
|
|
|
|
TcpServerConn::TcpServerConn(asio::ip::tcp::socket socket, TcpChannelCallbacks cbs)
|
|
|
|
|
: socket_(std::move(socket)),
|
|
|
|
|
strand_(asio::make_strand(socket_.get_executor())),
|
|
|
|
|
cbs_(std::move(cbs)) {}
|
|
|
|
|
|
|
|
|
|
TcpServerConn::TcpServerConn(asio::ip::tcp::socket socket, TcpChannelCallbacks cbs,
|
|
|
|
|
std::unique_ptr<crypto::TlsContext> tls)
|
|
|
|
|
: socket_(std::move(socket)),
|
|
|
|
|
strand_(asio::make_strand(socket_.get_executor())),
|
|
|
|
|
cbs_(std::move(cbs)),
|
|
|
|
|
tls_(std::move(tls)) {}
|
|
|
|
|
|
|
|
|
|
TcpServerConn::~TcpServerConn() {
|
|
|
|
|
close();
|
|
|
|
|
if (tls_thread_.joinable()) {
|
|
|
|
|
if (std::this_thread::get_id() == tls_thread_.get_id()) {
|
|
|
|
|
tls_thread_.detach(); // being destroyed from our own TLS thread — detach safely
|
|
|
|
|
} else {
|
|
|
|
|
tls_thread_.join();
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::start() {
|
|
|
|
|
if (tls_) {
|
|
|
|
|
// Run TLS handshake on a temporary thread so we don't block the io_context.
|
|
|
|
|
auto self = shared_from_this();
|
|
|
|
|
std::thread([self] {
|
|
|
|
|
std::string err;
|
|
|
|
|
int fd = static_cast<int>(self->socket_.native_handle());
|
|
|
|
|
if (!self->tls_->handshake(fd, err)) {
|
|
|
|
|
if (!self->closing_.exchange(true)) {
|
|
|
|
|
if (self->cbs_.on_error) {
|
|
|
|
|
asio::post(self->strand_, [self] {
|
|
|
|
|
self->cbs_.on_error(
|
|
|
|
|
std::make_error_code(std::errc::connection_reset));
|
|
|
|
|
});
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
self->connected_.store(true, std::memory_order_release);
|
feat(M2): UDP voice/media plane -- SFU relay, Opus, AEAD, jitter buffer
Adds the full voice pipeline: 14-byte binary frame header, ChaCha20-Poly1305
AEAD keyed from the TLS exporter, libopus encode/decode with FEC/PLC/DTX,
an adaptive per-ssrc jitter buffer, a miniaudio capture/playback engine, an
APM passthrough stub, and the UdpBinding/StreamAnnounce signaling chain
wired through ConnSession/SessionRegistry into a new server-side SFU
(MediaRelay) that decrypts and re-encrypts frames per channel member.
Exit criterion verified: test_m2_voice — two headless clients relay 50
encrypted Opus frames through the server; ctest --preset m1-dev is 9/9
green. Also corrects protocol.md's UdpBinding diagram, which described the
UDP-side binding packet as AEAD-sealed when it is in fact a plaintext
bootstrap frame (separate from the TCP/TLS UdpBinding ack).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-16 01:31:14 +02:00
|
|
|
// Export media keying material before the read loop starts.
|
|
|
|
|
if (self->cbs_.on_tls_ready) self->cbs_.on_tls_ready(*self->tls_);
|
feat(M1): TCP/TLS control plane -- auth, channels, ephemeral text
Implements the full M1 milestone. Two clients authenticate over TLS 1.3
(guest + Argon2id password) and exchange channel + private text messages
through a real server. All five ctest --preset m1-dev tests pass in ~1 s.
Key components added:
- vcpkg baseline + m1-dev preset (protobuf/mbedTLS/libsodium/asio/sqlite3)
- FrameCodec feed+emit, encode/decode_envelope, protobuf codegen
- TcpServerConn with blocking TLS handshake thread + tls_read_loop
- TlsContext (mbedTLS 1.3, ECDSA-P256 self-signed cert, TOFU on client)
- WorkerPool (3 threads, used for Argon2id)
- Database: SQLite + libsodium Argon2id, account lifecycle, bootstrap admin
- ServerIdentityManager: Ed25519 key + cert generate/persist/fingerprint
- ConnSession state machine: WaitingHello -> WaitingAuth -> Authenticated
- SessionRegistry: channel tree, user map, text routing, broadcast
- vc_client full M1 C ABI: connect/TLS/handshake/auth/text/disconnect
- voicecat-admin CLI: account add/reset/del/list
- test_m1_integration: M1 exit criterion, verified green
Bug fixed: double-framing in ConnSession::send_envelope -- encode_envelope
was adding the [4-byte len] prefix, then TcpServerConn::send_frame added
a second one, causing the client to parse [len][proto] as protobuf (silent
failure). Fixed by serializing raw protobuf bytes in send_envelope and
letting send_frame apply the single length prefix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-15 23:48:44 +02:00
|
|
|
// 50 ms timeout so tls_read_loop can drain the send queue between reads.
|
|
|
|
|
self->tls_->set_read_timeout(50);
|
|
|
|
|
self->tls_thread_ = std::thread([self] { self->tls_read_loop(); });
|
|
|
|
|
}).detach();
|
|
|
|
|
} else {
|
|
|
|
|
connected_.store(true, std::memory_order_release);
|
|
|
|
|
start_read();
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::tls_read_loop() {
|
|
|
|
|
std::vector<uint8_t> buf(16384);
|
|
|
|
|
while (!closing_.load(std::memory_order_acquire)) {
|
|
|
|
|
tls_drain_sends();
|
|
|
|
|
|
|
|
|
|
int n = tls_->read(buf.data(), buf.size());
|
|
|
|
|
if (crypto::TlsContext::is_timeout_error(n)) continue;
|
|
|
|
|
if (n <= 0) break;
|
|
|
|
|
|
|
|
|
|
std::vector<std::vector<uint8_t>> frames;
|
|
|
|
|
if (!codec_.feed(buf.data(), static_cast<size_t>(n), frames)) break;
|
|
|
|
|
for (auto& frame : frames) {
|
|
|
|
|
if (cbs_.on_frame) cbs_.on_frame(std::move(frame));
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
if (cbs_.on_disconnected) cbs_.on_disconnected();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::tls_drain_sends() {
|
|
|
|
|
while (true) {
|
|
|
|
|
std::vector<uint8_t> frame;
|
|
|
|
|
{
|
|
|
|
|
std::lock_guard lk(tls_send_mutex_);
|
|
|
|
|
if (tls_send_queue_.empty()) return;
|
|
|
|
|
frame = std::move(tls_send_queue_.front());
|
|
|
|
|
tls_send_queue_.pop_front();
|
|
|
|
|
}
|
|
|
|
|
size_t off = 0;
|
|
|
|
|
while (off < frame.size()) {
|
|
|
|
|
int n = tls_->write(frame.data() + off, frame.size() - off);
|
|
|
|
|
if (n <= 0) { closing_.store(true, std::memory_order_release); return; }
|
|
|
|
|
off += static_cast<size_t>(n);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::start_read() {
|
|
|
|
|
auto self = shared_from_this();
|
|
|
|
|
asio::async_read(socket_, asio::buffer(len_buf_, 4),
|
|
|
|
|
asio::bind_executor(strand_,
|
|
|
|
|
[this, self](std::error_code ec, std::size_t n) { handle_length(ec, n); }));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::handle_length(std::error_code ec, std::size_t) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
if (ec == asio::error::eof || ec == asio::error::connection_reset) {
|
|
|
|
|
if (cbs_.on_disconnected) cbs_.on_disconnected();
|
|
|
|
|
} else {
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(ec);
|
|
|
|
|
}
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
uint32_t length =
|
|
|
|
|
(static_cast<uint32_t>(len_buf_[0]) << 24) |
|
|
|
|
|
(static_cast<uint32_t>(len_buf_[1]) << 16) |
|
|
|
|
|
(static_cast<uint32_t>(len_buf_[2]) << 8) |
|
|
|
|
|
static_cast<uint32_t>(len_buf_[3]);
|
|
|
|
|
|
|
|
|
|
if (length > protocol::kMaxFrameBytes) {
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(asio::error::message_size);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
if (length == 0) {
|
|
|
|
|
if (cbs_.on_frame) cbs_.on_frame({});
|
|
|
|
|
start_read();
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
body_buf_.resize(length);
|
|
|
|
|
auto self = shared_from_this();
|
|
|
|
|
asio::async_read(socket_, asio::buffer(body_buf_),
|
|
|
|
|
asio::bind_executor(strand_,
|
|
|
|
|
[this, self, length](std::error_code ec, std::size_t n) {
|
|
|
|
|
handle_body(length, ec, n);
|
|
|
|
|
}));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::handle_body(uint32_t, std::error_code ec, std::size_t) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
if (ec == asio::error::eof || ec == asio::error::connection_reset) {
|
|
|
|
|
if (cbs_.on_disconnected) cbs_.on_disconnected();
|
|
|
|
|
} else {
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(ec);
|
|
|
|
|
}
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
if (cbs_.on_frame) cbs_.on_frame(body_buf_);
|
|
|
|
|
start_read();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::send_frame(std::vector<uint8_t> payload) {
|
|
|
|
|
std::vector<uint8_t> wire;
|
|
|
|
|
protocol::FrameCodec::emit(payload, wire);
|
|
|
|
|
if (tls_) {
|
|
|
|
|
std::lock_guard lk(tls_send_mutex_);
|
|
|
|
|
tls_send_queue_.push_back(std::move(wire));
|
|
|
|
|
} else {
|
|
|
|
|
auto self = shared_from_this();
|
|
|
|
|
asio::post(strand_, [this, self, w = std::move(wire)]() mutable {
|
|
|
|
|
send_queue_.push_back(std::move(w));
|
|
|
|
|
if (!sending_) do_send();
|
|
|
|
|
});
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::do_send() {
|
|
|
|
|
if (send_queue_.empty()) { sending_ = false; return; }
|
|
|
|
|
sending_ = true;
|
|
|
|
|
auto self = shared_from_this();
|
|
|
|
|
auto& front = send_queue_.front();
|
|
|
|
|
asio::async_write(socket_,
|
|
|
|
|
asio::buffer(front),
|
|
|
|
|
asio::bind_executor(strand_,
|
|
|
|
|
[this, self](std::error_code ec, std::size_t) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
if (cbs_.on_error) cbs_.on_error(ec);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
send_queue_.pop_front();
|
|
|
|
|
do_send();
|
|
|
|
|
}));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpServerConn::close() {
|
|
|
|
|
if (closing_.exchange(true)) return;
|
|
|
|
|
std::error_code ignored;
|
|
|
|
|
socket_.shutdown(asio::ip::tcp::socket::shutdown_both, ignored);
|
|
|
|
|
socket_.close(ignored);
|
|
|
|
|
connected_.store(false, std::memory_order_release);
|
|
|
|
|
// In non-TLS mode, the Asio async chain will naturally stop when the socket closes.
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// ── TcpAcceptor ─────────────────────────────────────────────────────────────
|
|
|
|
|
|
feat(M4): Windows WinForms client, TOFU identity pinning, VAD threshold + always-on mode
Core ABI extensions (voicecat.h):
- vc_list_channels / vc_list_users / vc_list_user_streams — pull-based snapshot getters
for the channel-tree and user-list UI; session_model_mu_ guards cross-thread reads
- VC_EVENT_JOIN_RESULT / vc_join_channel — channel join with optional password
- VC_EVENT_SERVER_IDENTITY + vc_confirm_server_identity — TOFU gate that blocks io_thread_
until the UI approves or rejects; pins TLS leaf-cert SHA-256 (not declared Ed25519)
- vc_get_server_identity_display — Ed25519 fingerprint for human-readable display only
- VC_INPUT_ALWAYS_ON = 2 in vc_input_mode — transmit unconditionally, no VAD gate
- vc_set_vad_threshold — live RMS threshold update (0.0–1.0); EnergyVadProcessor stores
it atomically so the audio RT path reads without a lock
C++ implementation:
- SessionModel::apply_snapshot / apply_channel_event fixed to populate parent_id,
password_protected, and max_users (were permanently zeroed)
- TlsContext::peer_cert_fingerprint — SHA-256 of peer leaf cert DER via mbedTLS
- TofuStore split into peek (read-only) + pin (write) so first-connect only persists
after user approval; tofu_store_path in vc_config for per-user pin file location
- TcpAcceptor uses dual-stack IPv6+IPv4 fallback (fixes localhost → ::1 on Windows)
- windows-client CMake preset: Release shared DLL, static MinGW runtime, no tools/tests
- New C++ tests: test_channel_user_list_abi, test_tofu_flow (14/14 green)
Windows client (clients/windows/ — .NET 10 WinForms):
- VoiceCat.Interop: LibraryImport P/Invoke surface, UnmanagedCallersOnly callbacks,
Channel<VoiceCatEvent> event delivery drained by 30ms WinForms Timer
- VoiceCat.App: ConnectDialog (saved servers, DPAPI password storage), ServerIdentity-
Dialog (TOFU first-connect / mismatch warning), MainForm (channel TreeView, user
ListBox, RichTextBox chat, voice controls, device pickers, VAD/PTT/always-on mode,
per-user gain/mute/NR tuning, VAD sensitivity TrackBar, level meter ProgressBar)
- PttKeyCaptureDialog — focus-scoped PTT key capture (documented limitation)
- PerUserTuningDialog — real-time gain/mute/NR applied to all of a user's streams
- Accessibility: explicit AccessibleName/Description on every control, & mnemonics,
Activity log ListBox as durable screen-reader record, AutomationNotification for
curated live announcements
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-17 00:35:16 +02:00
|
|
|
namespace {
|
|
|
|
|
// Try IPv6 dual-stack first (one socket handles both ::1 and 127.0.0.1 — fixes the common
|
|
|
|
|
// Windows case where `localhost` resolves to ::1 before 127.0.0.1). Falls back to IPv4-only
|
|
|
|
|
// if the OS has IPv6 disabled or the dual-stack bind fails for any reason.
|
|
|
|
|
asio::ip::tcp::acceptor make_acceptor(asio::io_context& io, uint16_t port) {
|
|
|
|
|
asio::ip::tcp::acceptor acc(io);
|
|
|
|
|
std::error_code ec;
|
|
|
|
|
acc.open(asio::ip::tcp::v6(), ec);
|
|
|
|
|
if (!ec) {
|
|
|
|
|
acc.set_option(asio::ip::v6_only(false), ec); // dual-stack
|
|
|
|
|
acc.set_option(asio::ip::tcp::acceptor::reuse_address(true));
|
|
|
|
|
acc.bind(asio::ip::tcp::endpoint(asio::ip::tcp::v6(), port), ec);
|
|
|
|
|
if (!ec) acc.listen(asio::socket_base::max_listen_connections, ec);
|
|
|
|
|
}
|
|
|
|
|
if (ec) {
|
|
|
|
|
if (acc.is_open()) { std::error_code ignored; acc.close(ignored); }
|
|
|
|
|
acc.open(asio::ip::tcp::v4());
|
|
|
|
|
acc.set_option(asio::ip::tcp::acceptor::reuse_address(true));
|
|
|
|
|
acc.bind(asio::ip::tcp::endpoint(asio::ip::tcp::v4(), port));
|
|
|
|
|
acc.listen(asio::socket_base::max_listen_connections);
|
|
|
|
|
}
|
|
|
|
|
return acc;
|
feat(M1): TCP/TLS control plane -- auth, channels, ephemeral text
Implements the full M1 milestone. Two clients authenticate over TLS 1.3
(guest + Argon2id password) and exchange channel + private text messages
through a real server. All five ctest --preset m1-dev tests pass in ~1 s.
Key components added:
- vcpkg baseline + m1-dev preset (protobuf/mbedTLS/libsodium/asio/sqlite3)
- FrameCodec feed+emit, encode/decode_envelope, protobuf codegen
- TcpServerConn with blocking TLS handshake thread + tls_read_loop
- TlsContext (mbedTLS 1.3, ECDSA-P256 self-signed cert, TOFU on client)
- WorkerPool (3 threads, used for Argon2id)
- Database: SQLite + libsodium Argon2id, account lifecycle, bootstrap admin
- ServerIdentityManager: Ed25519 key + cert generate/persist/fingerprint
- ConnSession state machine: WaitingHello -> WaitingAuth -> Authenticated
- SessionRegistry: channel tree, user map, text routing, broadcast
- vc_client full M1 C ABI: connect/TLS/handshake/auth/text/disconnect
- voicecat-admin CLI: account add/reset/del/list
- test_m1_integration: M1 exit criterion, verified green
Bug fixed: double-framing in ConnSession::send_envelope -- encode_envelope
was adding the [4-byte len] prefix, then TcpServerConn::send_frame added
a second one, causing the client to parse [len][proto] as protobuf (silent
failure). Fixed by serializing raw protobuf bytes in send_envelope and
letting send_frame apply the single length prefix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-15 23:48:44 +02:00
|
|
|
}
|
feat(M4): Windows WinForms client, TOFU identity pinning, VAD threshold + always-on mode
Core ABI extensions (voicecat.h):
- vc_list_channels / vc_list_users / vc_list_user_streams — pull-based snapshot getters
for the channel-tree and user-list UI; session_model_mu_ guards cross-thread reads
- VC_EVENT_JOIN_RESULT / vc_join_channel — channel join with optional password
- VC_EVENT_SERVER_IDENTITY + vc_confirm_server_identity — TOFU gate that blocks io_thread_
until the UI approves or rejects; pins TLS leaf-cert SHA-256 (not declared Ed25519)
- vc_get_server_identity_display — Ed25519 fingerprint for human-readable display only
- VC_INPUT_ALWAYS_ON = 2 in vc_input_mode — transmit unconditionally, no VAD gate
- vc_set_vad_threshold — live RMS threshold update (0.0–1.0); EnergyVadProcessor stores
it atomically so the audio RT path reads without a lock
C++ implementation:
- SessionModel::apply_snapshot / apply_channel_event fixed to populate parent_id,
password_protected, and max_users (were permanently zeroed)
- TlsContext::peer_cert_fingerprint — SHA-256 of peer leaf cert DER via mbedTLS
- TofuStore split into peek (read-only) + pin (write) so first-connect only persists
after user approval; tofu_store_path in vc_config for per-user pin file location
- TcpAcceptor uses dual-stack IPv6+IPv4 fallback (fixes localhost → ::1 on Windows)
- windows-client CMake preset: Release shared DLL, static MinGW runtime, no tools/tests
- New C++ tests: test_channel_user_list_abi, test_tofu_flow (14/14 green)
Windows client (clients/windows/ — .NET 10 WinForms):
- VoiceCat.Interop: LibraryImport P/Invoke surface, UnmanagedCallersOnly callbacks,
Channel<VoiceCatEvent> event delivery drained by 30ms WinForms Timer
- VoiceCat.App: ConnectDialog (saved servers, DPAPI password storage), ServerIdentity-
Dialog (TOFU first-connect / mismatch warning), MainForm (channel TreeView, user
ListBox, RichTextBox chat, voice controls, device pickers, VAD/PTT/always-on mode,
per-user gain/mute/NR tuning, VAD sensitivity TrackBar, level meter ProgressBar)
- PttKeyCaptureDialog — focus-scoped PTT key capture (documented limitation)
- PerUserTuningDialog — real-time gain/mute/NR applied to all of a user's streams
- Accessibility: explicit AccessibleName/Description on every control, & mnemonics,
Activity log ListBox as durable screen-reader record, AutomationNotification for
curated live announcements
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-17 00:35:16 +02:00
|
|
|
} // namespace
|
|
|
|
|
|
|
|
|
|
TcpAcceptor::TcpAcceptor(asio::io_context& io, uint16_t port, ConnFactory factory)
|
|
|
|
|
: acceptor_(make_acceptor(io, port)),
|
|
|
|
|
factory_(std::move(factory)) {}
|
feat(M1): TCP/TLS control plane -- auth, channels, ephemeral text
Implements the full M1 milestone. Two clients authenticate over TLS 1.3
(guest + Argon2id password) and exchange channel + private text messages
through a real server. All five ctest --preset m1-dev tests pass in ~1 s.
Key components added:
- vcpkg baseline + m1-dev preset (protobuf/mbedTLS/libsodium/asio/sqlite3)
- FrameCodec feed+emit, encode/decode_envelope, protobuf codegen
- TcpServerConn with blocking TLS handshake thread + tls_read_loop
- TlsContext (mbedTLS 1.3, ECDSA-P256 self-signed cert, TOFU on client)
- WorkerPool (3 threads, used for Argon2id)
- Database: SQLite + libsodium Argon2id, account lifecycle, bootstrap admin
- ServerIdentityManager: Ed25519 key + cert generate/persist/fingerprint
- ConnSession state machine: WaitingHello -> WaitingAuth -> Authenticated
- SessionRegistry: channel tree, user map, text routing, broadcast
- vc_client full M1 C ABI: connect/TLS/handshake/auth/text/disconnect
- voicecat-admin CLI: account add/reset/del/list
- test_m1_integration: M1 exit criterion, verified green
Bug fixed: double-framing in ConnSession::send_envelope -- encode_envelope
was adding the [4-byte len] prefix, then TcpServerConn::send_frame added
a second one, causing the client to parse [len][proto] as protobuf (silent
failure). Fixed by serializing raw protobuf bytes in send_envelope and
letting send_frame apply the single length prefix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-15 23:48:44 +02:00
|
|
|
|
|
|
|
|
void TcpAcceptor::start() { do_accept(); }
|
|
|
|
|
|
|
|
|
|
void TcpAcceptor::stop() {
|
|
|
|
|
stopped_ = true;
|
|
|
|
|
std::error_code ignored;
|
|
|
|
|
acceptor_.close(ignored);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void TcpAcceptor::do_accept() {
|
|
|
|
|
if (stopped_) return;
|
|
|
|
|
acceptor_.async_accept(
|
|
|
|
|
[this](std::error_code ec, asio::ip::tcp::socket socket) {
|
|
|
|
|
if (ec) {
|
|
|
|
|
if (!stopped_) do_accept();
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
socket.set_option(asio::ip::tcp::no_delay(true));
|
|
|
|
|
auto conn = factory_(std::move(socket));
|
|
|
|
|
if (conn) conn->start();
|
|
|
|
|
do_accept();
|
|
|
|
|
});
|
|
|
|
|
}
|
scaffold: M0 skeleton + agent onboarding (build, architecture, progress)
Turn the design into a buildable, dependency-free M0 skeleton plus the
onboarding layer so a new agent can pick up instantly.
Build system:
- CMake + CMakePresets (dev = no deps; server-release = vcpkg) + vcpkg.json
- Skeleton builds with just a C++20 compiler; deps stay off until needed
- .gitattributes (LF), .gitignore, .clang-format
Core (libvoicecat):
- core/include/voicecat.h: full C ABI (the client/server contract), stubbed
- core/proto/voicecat.proto: control-plane wire format, matches docs/protocol.md
- src/{net,crypto,codec,protocol,session,audio,core}: subsystem stubs that
return VC_ERR_NOT_IMPLEMENTED, each pointing to its design doc
- server/ (voicecat-server) and tools/vccli/ link the core
- tests/: CTest smoke test asserting the C ABI contract (behavior, not just build)
- clients/{apple,windows}: M4 placeholders
Onboarding for agents:
- CLAUDE.md: hub — build/test commands, architecture at a glance, doc map, rules
- AGENTS.md: working method (behavior-driven; clean compile is the floor not the goal)
- PROGRESS.md: living tracker — M0 done, M1 task checklist, "where we left off"
Verified: cmake --preset dev && cmake --build --preset dev && ctest --preset dev → green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-15 21:09:09 +02:00
|
|
|
|
feat(M2): UDP voice/media plane -- SFU relay, Opus, AEAD, jitter buffer
Adds the full voice pipeline: 14-byte binary frame header, ChaCha20-Poly1305
AEAD keyed from the TLS exporter, libopus encode/decode with FEC/PLC/DTX,
an adaptive per-ssrc jitter buffer, a miniaudio capture/playback engine, an
APM passthrough stub, and the UdpBinding/StreamAnnounce signaling chain
wired through ConnSession/SessionRegistry into a new server-side SFU
(MediaRelay) that decrypts and re-encrypts frames per channel member.
Exit criterion verified: test_m2_voice — two headless clients relay 50
encrypted Opus frames through the server; ctest --preset m1-dev is 9/9
green. Also corrects protocol.md's UdpBinding diagram, which described the
UDP-side binding packet as AEAD-sealed when it is in fact a plaintext
bootstrap frame (separate from the TCP/TLS UdpBinding ack).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-16 01:31:14 +02:00
|
|
|
// ── UdpMediaChannel ──────────────────────────────────────────────────────────
|
|
|
|
|
|
|
|
|
|
bool UdpMediaChannel::bind(asio::io_context& io, uint16_t port) {
|
|
|
|
|
if (bound_.load()) return false;
|
|
|
|
|
try {
|
|
|
|
|
socket_ = std::make_unique<asio::ip::udp::socket>(io);
|
|
|
|
|
socket_->open(asio::ip::udp::v4());
|
|
|
|
|
socket_->set_option(asio::socket_base::reuse_address(true));
|
|
|
|
|
socket_->bind(asio::ip::udp::endpoint(asio::ip::udp::v4(), port));
|
|
|
|
|
bound_.store(true, std::memory_order_release);
|
|
|
|
|
return true;
|
|
|
|
|
} catch (...) {
|
|
|
|
|
socket_.reset();
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void UdpMediaChannel::start_recv(FrameCallback cb) {
|
|
|
|
|
frame_cb_ = std::move(cb);
|
|
|
|
|
do_recv();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void UdpMediaChannel::do_recv() {
|
|
|
|
|
if (!socket_ || closed_.load()) return;
|
|
|
|
|
socket_->async_receive_from(
|
|
|
|
|
asio::buffer(recv_buf_), sender_ep_,
|
|
|
|
|
[this](std::error_code ec, std::size_t n) {
|
|
|
|
|
if (ec || closed_.load()) return;
|
|
|
|
|
if (frame_cb_ && n > 0)
|
|
|
|
|
frame_cb_(recv_buf_.data(), n, sender_ep_);
|
|
|
|
|
do_recv();
|
|
|
|
|
});
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void UdpMediaChannel::send_to(const uint8_t* data, size_t len,
|
|
|
|
|
asio::ip::udp::endpoint dst) {
|
|
|
|
|
if (!socket_ || closed_.load() || len == 0) return;
|
|
|
|
|
auto buf = std::make_shared<std::vector<uint8_t>>(data, data + len);
|
|
|
|
|
asio::post(socket_->get_executor(), [this, buf, dst]() mutable {
|
|
|
|
|
if (closed_.load()) return;
|
|
|
|
|
socket_->async_send_to(
|
|
|
|
|
asio::buffer(*buf), dst,
|
|
|
|
|
[buf](std::error_code, std::size_t) {});
|
|
|
|
|
});
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void UdpMediaChannel::close() {
|
|
|
|
|
if (closed_.exchange(true)) return;
|
|
|
|
|
if (socket_) {
|
|
|
|
|
std::error_code ec;
|
|
|
|
|
socket_->cancel(ec);
|
|
|
|
|
socket_->close(ec);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
asio::ip::udp::endpoint UdpMediaChannel::local_endpoint() const {
|
|
|
|
|
if (!socket_) return {};
|
|
|
|
|
std::error_code ec;
|
|
|
|
|
return socket_->local_endpoint(ec);
|
|
|
|
|
}
|
|
|
|
|
|
scaffold: M0 skeleton + agent onboarding (build, architecture, progress)
Turn the design into a buildable, dependency-free M0 skeleton plus the
onboarding layer so a new agent can pick up instantly.
Build system:
- CMake + CMakePresets (dev = no deps; server-release = vcpkg) + vcpkg.json
- Skeleton builds with just a C++20 compiler; deps stay off until needed
- .gitattributes (LF), .gitignore, .clang-format
Core (libvoicecat):
- core/include/voicecat.h: full C ABI (the client/server contract), stubbed
- core/proto/voicecat.proto: control-plane wire format, matches docs/protocol.md
- src/{net,crypto,codec,protocol,session,audio,core}: subsystem stubs that
return VC_ERR_NOT_IMPLEMENTED, each pointing to its design doc
- server/ (voicecat-server) and tools/vccli/ link the core
- tests/: CTest smoke test asserting the C ABI contract (behavior, not just build)
- clients/{apple,windows}: M4 placeholders
Onboarding for agents:
- CLAUDE.md: hub — build/test commands, architecture at a glance, doc map, rules
- AGENTS.md: working method (behavior-driven; clean compile is the floor not the goal)
- PROGRESS.md: living tracker — M0 done, M1 task checklist, "where we left off"
Verified: cmake --preset dev && cmake --build --preset dev && ctest --preset dev → green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-15 21:09:09 +02:00
|
|
|
} // namespace voicecat::net
|
feat(M1): TCP/TLS control plane -- auth, channels, ephemeral text
Implements the full M1 milestone. Two clients authenticate over TLS 1.3
(guest + Argon2id password) and exchange channel + private text messages
through a real server. All five ctest --preset m1-dev tests pass in ~1 s.
Key components added:
- vcpkg baseline + m1-dev preset (protobuf/mbedTLS/libsodium/asio/sqlite3)
- FrameCodec feed+emit, encode/decode_envelope, protobuf codegen
- TcpServerConn with blocking TLS handshake thread + tls_read_loop
- TlsContext (mbedTLS 1.3, ECDSA-P256 self-signed cert, TOFU on client)
- WorkerPool (3 threads, used for Argon2id)
- Database: SQLite + libsodium Argon2id, account lifecycle, bootstrap admin
- ServerIdentityManager: Ed25519 key + cert generate/persist/fingerprint
- ConnSession state machine: WaitingHello -> WaitingAuth -> Authenticated
- SessionRegistry: channel tree, user map, text routing, broadcast
- vc_client full M1 C ABI: connect/TLS/handshake/auth/text/disconnect
- voicecat-admin CLI: account add/reset/del/list
- test_m1_integration: M1 exit criterion, verified green
Bug fixed: double-framing in ConnSession::send_envelope -- encode_envelope
was adding the [4-byte len] prefix, then TcpServerConn::send_frame added
a second one, causing the client to parse [len][proto] as protobuf (silent
failure). Fixed by serializing raw protobuf bytes in send_envelope and
letting send_frame apply the single length prefix.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-06-15 23:48:44 +02:00
|
|
|
|
|
|
|
|
#endif // VOICECAT_HAS_NET
|