Add mechanism for recovering from disconnections in conferences

* add freezing and unfreezing of peers
  * add rejoin packet
  * revise handling of temporary invited connections
  * rename "peer kill" packet to "peer leave" packet
  * test rejoining in conference test
  * use custom clock in conference test

auto_tests/conference_double_invite_test.c

@@ -81,6 +81,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(2, conference_double_invite_test);
+    run_auto_test(2, conference_double_invite_test, false);
     return 0;
 }

auto_tests/conference_peer_nick_test.c

@@ -129,6 +129,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(2, conference_peer_nick_test);
+    run_auto_test(2, conference_peer_nick_test, false);
     return 0;
 }

auto_tests/conference_test.c

@@ -16,26 +16,31 @@
 #include "check_compat.h"
 
 #define NUM_GROUP_TOX 16
+#define NUM_DISCONNECT 8
 #define GROUP_MESSAGE "Install Gentoo"
 
-#define NAME_FORMAT_STR "Tox #%4u"
 #define NAMELEN 9
-#define NAME_FORMAT "%9s"
+#define NAME_FORMAT_STR "Tox #%4u"
+#define NEW_NAME_FORMAT_STR "New #%4u"
 
 typedef struct State {
-    uint32_t id;
+    uint32_t index;
+    uint64_t clock;
+
     bool invited_next;
 } State;
 
+#include "run_auto_test.h"
+
 static void handle_self_connection_status(
     Tox *tox, TOX_CONNECTION connection_status, void *user_data)
 {
     const State *state = (State *)user_data;
 
     if (connection_status != TOX_CONNECTION_NONE) {
-        printf("tox #%u: is now connected\n", state->id);
+        printf("tox #%u: is now connected\n", state->index);
     } else {
-        printf("tox #%u: is now disconnected\n", state->id);
+        printf("tox #%u: is now disconnected\n", state->index);
     }
 }
 
@@ -45,9 +50,9 @@ static void handle_friend_connection_status(
     const State *state = (State *)user_data;
 
     if (connection_status != TOX_CONNECTION_NONE) {
-        printf("tox #%u: is now connected to friend %u\n", state->id, friendnumber);
+        printf("tox #%u: is now connected to friend %u\n", state->index, friendnumber);
     } else {
-        printf("tox #%u: is now disconnected from friend %u\n", state->id, friendnumber);
+        printf("tox #%u: is now disconnected from friend %u\n", state->index, friendnumber);
     }
 }
 
@@ -56,18 +61,18 @@ static void handle_conference_invite(
     const uint8_t *data, size_t length, void *user_data)
 {
     const State *state = (State *)user_data;
-    ck_assert_msg(type == TOX_CONFERENCE_TYPE_TEXT, "tox #%u: wrong conference type: %d", state->id, type);
+    ck_assert_msg(type == TOX_CONFERENCE_TYPE_TEXT, "tox #%u: wrong conference type: %d", state->index, type);
 
     TOX_ERR_CONFERENCE_JOIN err;
     uint32_t g_num = tox_conference_join(tox, friendnumber, data, length, &err);
 
-    ck_assert_msg(err == TOX_ERR_CONFERENCE_JOIN_OK, "tox #%u: error joining group: %d", state->id, err);
-    ck_assert_msg(g_num == 0, "tox #%u: group number was not 0", state->id);
+    ck_assert_msg(err == TOX_ERR_CONFERENCE_JOIN_OK, "tox #%u: error joining group: %d", state->index, err);
+    ck_assert_msg(g_num == 0, "tox #%u: group number was not 0", state->index);
 
     // Try joining again. We should only be allowed to join once.
     tox_conference_join(tox, friendnumber, data, length, &err);
     ck_assert_msg(err != TOX_ERR_CONFERENCE_JOIN_OK,
-                  "tox #%u: joining groupchat twice should be impossible.", state->id);
+                  "tox #%u: joining groupchat twice should be impossible.", state->index);
 }
 
 static void handle_conference_connected(
@@ -81,8 +86,8 @@ static void handle_conference_connected(
 
     TOX_ERR_CONFERENCE_INVITE err;
     tox_conference_invite(tox, 1, 0, &err);
-    ck_assert_msg(err == TOX_ERR_CONFERENCE_INVITE_OK, "tox #%u failed to invite next friend: err = %d", state->id, err);
-    printf("tox #%u: invited next friend\n", state->id);
+    ck_assert_msg(err == TOX_ERR_CONFERENCE_INVITE_OK, "tox #%u failed to invite next friend: err = %d", state->index, err);
+    printf("tox #%u: invited next friend\n", state->index);
     state->invited_next = true;
 }
 
@@ -97,8 +102,93 @@ static void handle_conference_message(
     }
 }
 
+static bool toxes_are_disconnected_from_group(uint32_t tox_count, Tox **toxes, int disconnected_count,
+        bool *disconnected)
+{
+    for (uint32_t i = 0; i < tox_count; i++) {
+        if (disconnected[i]) {
+            continue;
+        }
+
+        if (tox_conference_peer_count(toxes[i], 0, nullptr) > tox_count - NUM_DISCONNECT) {
+            return false;
+        }
+    }
+
+    return true;
+}
+
+static bool all_connected_to_group(uint32_t tox_count, Tox **toxes)
+{
+    for (uint32_t i = 0; i < tox_count; i++) {
+        if (tox_conference_peer_count(toxes[i], 0, nullptr) < tox_count) {
+            return false;
+        }
+    }
+
+    return true;
+}
+
+/* returns a random index at which a list of booleans is false
+ * (some such index is required to exist)
+ * */
+static uint32_t random_false_index(bool *list, const uint32_t length)
+{
+    uint32_t index;
+
+    do {
+        index = random_u32() % length;
+    } while (list[index]);
+
+    return index;
+}
+
 static void run_conference_tests(Tox **toxes, State *state)
 {
+    /* disabling name propagation check for now, as it occasionally fails due
+     * to disconnections too short to trigger freezing */
+    const bool check_name_propagation = false;
+
+    printf("letting random toxes timeout\n");
+    bool disconnected[NUM_GROUP_TOX] = {0};
+
+    ck_assert(NUM_DISCONNECT < NUM_GROUP_TOX);
+
+    for (uint16_t i = 0; i < NUM_DISCONNECT; ++i) {
+        uint32_t disconnect = random_false_index(disconnected, NUM_GROUP_TOX);
+        disconnected[disconnect] = true;
+        printf("Disconnecting #%u\n", state[disconnect].index);
+    }
+
+    do {
+        for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
+            if (!disconnected[i]) {
+                tox_iterate(toxes[i], &state[i]);
+                state[i].clock += 1000;
+            }
+        }
+
+        c_sleep(20);
+    } while (!toxes_are_disconnected_from_group(NUM_GROUP_TOX, toxes, NUM_DISCONNECT, disconnected));
+
+    if (check_name_propagation) {
+        printf("changing names\n");
+
+        for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
+            char name[NAMELEN + 1];
+            snprintf(name, NAMELEN + 1, NEW_NAME_FORMAT_STR, state[i].index);
+            tox_self_set_name(toxes[i], (const uint8_t *)name, NAMELEN, nullptr);
+        }
+    }
+
+    printf("reconnecting toxes\n");
+
+    do {
+        iterate_all_wait(NUM_GROUP_TOX, toxes, state, ITERATION_INTERVAL);
+    } while (!all_connected_to_group(NUM_GROUP_TOX, toxes));
+
+    printf("running conference tests\n");
+
     for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
         tox_callback_conference_message(toxes[i], &handle_conference_message);
     }
@@ -112,41 +202,33 @@ static void run_conference_tests(Tox **toxes, State *state)
         err == TOX_ERR_CONFERENCE_SEND_MESSAGE_OK, "failed to send group message");
     num_recv = 0;
 
-    for (uint8_t j = 0; j < 20; ++j) {
-        for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
-            tox_iterate(toxes[i], &state[i]);
+    for (uint8_t j = 0; j < NUM_GROUP_TOX * 2; ++j) {
+        iterate_all_wait(NUM_GROUP_TOX, toxes, state, ITERATION_INTERVAL);
     }
 
-        c_sleep(25);
-    }
-
-    c_sleep(25);
     ck_assert_msg(num_recv == NUM_GROUP_TOX, "failed to recv group messages");
 
     for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
         for (uint16_t j = 0; j < NUM_GROUP_TOX; ++j) {
             const size_t len = tox_conference_peer_get_name_size(toxes[i], 0, j, nullptr);
-            ck_assert_msg(len == NAMELEN, "name of #%u according to #%u has incorrect length %u", state[j].id, state[i].id,
+            ck_assert_msg(len == NAMELEN, "name of #%u according to #%u has incorrect length %u", state[j].index, state[i].index,
                           (unsigned int)len);
+
+            if (check_name_propagation) {
                 uint8_t name[NAMELEN];
                 tox_conference_peer_get_name(toxes[i], 0, j, name, nullptr);
-            char expected_name[NAMELEN + 1];
-            snprintf(expected_name, NAMELEN + 1, NAME_FORMAT_STR, state[j].id);
-            ck_assert_msg(memcmp(name, expected_name, NAMELEN) == 0,
-                          "name of #%u according to #%u is \"" NAME_FORMAT "\"; expected \"%s\"",
-                          state[j].id, state[i].id, name, expected_name);
+                /* Note the toxes will have been reordered */
+                ck_assert_msg(memcmp(name, "New", 3) == 0,
+                              "name of #%u according to #%u not updated", state[j].index, state[i].index);
+            }
         }
     }
 
     for (uint16_t k = NUM_GROUP_TOX; k != 0 ; --k) {
         tox_conference_delete(toxes[k - 1], 0, nullptr);
 
-        for (uint8_t j = 0; j < 10; ++j) {
-            for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
-                tox_iterate(toxes[i], &state[i]);
-            }
-
-            c_sleep(50);
+        for (uint8_t j = 0; j < 10 || j < NUM_GROUP_TOX; ++j) {
+            iterate_all_wait(NUM_GROUP_TOX, toxes, state, ITERATION_INTERVAL);
         }
 
         for (uint16_t i = 0; i < k - 1; ++i) {
@@ -158,82 +240,23 @@ static void run_conference_tests(Tox **toxes, State *state)
     }
 }
 
-static void test_many_group(void)
+static void test_many_group(Tox **toxes, State *state)
 {
     const time_t test_start_time = time(nullptr);
 
-    Tox *toxes[NUM_GROUP_TOX];
-    State state[NUM_GROUP_TOX];
-    memset(state, 0, NUM_GROUP_TOX * sizeof(State));
-    time_t cur_time = time(nullptr);
-    struct Tox_Options *opts = tox_options_new(nullptr);
-    tox_options_set_start_port(opts, 33445);
-    tox_options_set_end_port(opts, 34445);
-
-    printf("creating %d toxes\n", NUM_GROUP_TOX);
-
     for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
-        TOX_ERR_NEW err;
-        state[i].id = i + 1;
-        toxes[i] = tox_new_log(opts, &err, &state[i]);
-
-        ck_assert_msg(toxes[i] != nullptr, "failed to create tox instance %u: error %d", i, err);
         tox_callback_self_connection_status(toxes[i], &handle_self_connection_status);
         tox_callback_friend_connection_status(toxes[i], &handle_friend_connection_status);
         tox_callback_conference_invite(toxes[i], &handle_conference_invite);
         tox_callback_conference_connected(toxes[i], &handle_conference_connected);
 
         char name[NAMELEN + 1];
-        snprintf(name, NAMELEN + 1, NAME_FORMAT_STR, state[i].id);
+        snprintf(name, NAMELEN + 1, NAME_FORMAT_STR, state[i].index);
         tox_self_set_name(toxes[i], (const uint8_t *)name, NAMELEN, nullptr);
-
-        if (i != 0) {
-            uint8_t dht_key[TOX_PUBLIC_KEY_SIZE];
-            tox_self_get_dht_id(toxes[0], dht_key);
-            const uint16_t dht_port = tox_self_get_udp_port(toxes[0], nullptr);
-
-            tox_bootstrap(toxes[i], "localhost", dht_port, dht_key, nullptr);
     }
-    }
-
-    tox_options_free(opts);
-
-    printf("creating a chain of friends\n");
-
-    for (unsigned i = 1; i < NUM_GROUP_TOX; ++i) {
-        TOX_ERR_FRIEND_ADD err;
-        uint8_t key[TOX_PUBLIC_KEY_SIZE];
-
-        tox_self_get_public_key(toxes[i - 1], key);
-        tox_friend_add_norequest(toxes[i], key, &err);
-        ck_assert_msg(err == TOX_ERR_FRIEND_ADD_OK, "failed to add friend: error %d", err);
-
-        tox_self_get_public_key(toxes[i], key);
-        tox_friend_add_norequest(toxes[i - 1], key, &err);
-        ck_assert_msg(err == TOX_ERR_FRIEND_ADD_OK, "failed to add friend: error %d", err);
-    }
-
-    printf("waiting for everyone to come online\n");
-    unsigned online_count = 0;
-
-    do {
-        online_count = 0;
-
-        for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
-            tox_iterate(toxes[i], &state[i]);
-            online_count += tox_friend_get_connection_status(toxes[i], 0, nullptr) != TOX_CONNECTION_NONE;
-        }
-
-        printf("currently %u toxes are online\n", online_count);
-        fflush(stdout);
-
-        c_sleep(1000);
-    } while (online_count != NUM_GROUP_TOX);
-
-    printf("friends connected, took %d seconds\n", (int)(time(nullptr) - cur_time));
 
     ck_assert_msg(tox_conference_new(toxes[0], nullptr) != UINT32_MAX, "failed to create group");
-    printf("tox #%u: inviting its first friend\n", state[0].id);
+    printf("tox #%u: inviting its first friend\n", state[0].index);
     ck_assert_msg(tox_conference_invite(toxes[0], 0, 0, nullptr) != 0, "failed to invite friend");
     state[0].invited_next = true;
     ck_assert_msg(tox_conference_set_title(toxes[0], 0, (const uint8_t *)"Gentoo", sizeof("Gentoo") - 1, nullptr) != 0,
@@ -244,17 +267,16 @@ static void test_many_group(void)
     uint16_t invited_count = 0;
 
     do {
+        iterate_all_wait(NUM_GROUP_TOX, toxes, state, ITERATION_INTERVAL);
+
         invited_count = 0;
 
         for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
-            tox_iterate(toxes[i], &state[i]);
             invited_count += state[i].invited_next;
         }
-
-        c_sleep(50);
     } while (invited_count != NUM_GROUP_TOX - 1);
 
-    cur_time = time(nullptr);
+    uint64_t pregroup_clock = state[0].clock;
     printf("waiting for all toxes to be in the group\n");
     uint16_t fully_connected_count = 0;
 
@@ -262,8 +284,9 @@ static void test_many_group(void)
         fully_connected_count = 0;
         printf("current peer counts: [");
 
+        iterate_all_wait(NUM_GROUP_TOX, toxes, state, ITERATION_INTERVAL);
+
         for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
-            tox_iterate(toxes[i], &state[i]);
             TOX_ERR_CONFERENCE_PEER_QUERY err;
             uint32_t peer_count = tox_conference_peer_count(toxes[i], 0, &err);
 
@@ -282,8 +305,6 @@ static void test_many_group(void)
 
         printf("]\n");
         fflush(stdout);
-
-        c_sleep(200);
     } while (fully_connected_count != NUM_GROUP_TOX);
 
     for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
@@ -300,16 +321,10 @@ static void test_many_group(void)
         ck_assert_msg(memcmp("Gentoo", title, ret) == 0, "Wrong title");
     }
 
-    printf("group connected, took %d seconds\n", (int)(time(nullptr) - cur_time));
+    printf("group connected, took %d seconds\n", (int)((state[0].clock - pregroup_clock) / 1000));
 
     run_conference_tests(toxes, state);
 
-    printf("tearing down toxes\n");
-
-    for (uint16_t i = 0; i < NUM_GROUP_TOX; ++i) {
-        tox_kill(toxes[i]);
-    }
-
     printf("test_many_group succeeded, took %d seconds\n", (int)(time(nullptr) - test_start_time));
 }
 
@@ -317,6 +332,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    test_many_group();
+    run_auto_test(NUM_GROUP_TOX, test_many_group, true);
     return 0;
 }

auto_tests/friend_connection_test.c

@@ -27,6 +27,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(2, friend_connection_test);
+    run_auto_test(2, friend_connection_test, false);
     return 0;
 }

auto_tests/lossless_packet_test.c

@@ -61,6 +61,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(2, test_lossless_packet);
+    run_auto_test(2, test_lossless_packet, false);
     return 0;
 }

auto_tests/lossy_packet_test.c

@@ -57,6 +57,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(2, test_lossy_packet);
+    run_auto_test(2, test_lossy_packet, false);
     return 0;
 }

auto_tests/overflow_recvq_test.c

@@ -62,6 +62,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(3, net_crypto_overflow_test);
+    run_auto_test(3, net_crypto_overflow_test, false);
     return 0;
 }

auto_tests/overflow_sendq_test.c

@@ -47,6 +47,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(2, net_crypto_overflow_test);
+    run_auto_test(2, net_crypto_overflow_test, false);
     return 0;
 }

auto_tests/reconnect_test.c

@@ -101,6 +101,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(TOX_COUNT, test_reconnect);
+    run_auto_test(TOX_COUNT, test_reconnect, false);
     return 0;
 }

auto_tests/run_auto_test.h

@@ -48,7 +48,7 @@ static uint64_t get_state_clock_callback(Mono_Time *mono_time, void *user_data)
     return state->clock;
 }
 
-static void run_auto_test(uint32_t tox_count, void test(Tox **toxes, State *state))
+static void run_auto_test(uint32_t tox_count, void test(Tox **toxes, State *state), bool chain)
 {
     printf("initialising %u toxes\n", tox_count);
     Tox **toxes = (Tox **)calloc(tox_count, sizeof(Tox *));
@@ -66,6 +66,21 @@ static void run_auto_test(uint32_t tox_count, void test(Tox **toxes, State *stat
         mono_time_set_current_time_callback(mono_time, get_state_clock_callback, &state[i]);
     }
 
+    if (chain) {
+        printf("each tox adds adjacent toxes as friends\n");
+
+        for (uint32_t i = 0; i < tox_count; i++) {
+            for (uint32_t j = i - 1; j != i + 3; j += 2) {
+                if (j >= tox_count) {
+                    continue;
+                }
+
+                uint8_t public_key[TOX_PUBLIC_KEY_SIZE];
+                tox_self_get_public_key(toxes[j], public_key);
+                tox_friend_add_norequest(toxes[i], public_key, nullptr);
+            }
+        }
+    } else {
         printf("toxes all add each other as friends\n");
 
         for (uint32_t i = 0; i < tox_count; i++) {
@@ -77,7 +92,7 @@ static void run_auto_test(uint32_t tox_count, void test(Tox **toxes, State *stat
                 }
             }
         }
-
+    }
 
     printf("bootstrapping all toxes off toxes[0]\n");
     uint8_t dht_key[TOX_PUBLIC_KEY_SIZE];

auto_tests/send_message_test.c

@@ -61,6 +61,6 @@ int main(void)
 {
     setvbuf(stdout, nullptr, _IONBF, 0);
 
-    run_auto_test(2, send_message_test);
+    run_auto_test(2, send_message_test, false);
     return 0;
 }

docs/minpgc.md

@@ -0,0 +1,128 @@
+# Persistent conferences
+
+This document describes the "minpgc" simple persistent conferences
+implementation of PR #1069.
+
+Many of the ideas derive from isotoxin's persistent conferences
+implementation, PR #826.
+
+## Specification of changes from pre-existing conference specification
+We add one new packet type:
+
+Rejoin Conference packet
+
+| Length | Contents                        |
+|:-------|:--------------------------------|
+| `1`    | `uint8_t` (0x64)                |
+| `33`   | Group chat identifier           |
+
+
+A peer times out from a group if it has been inactive for 60s. When a peer
+times out, we flag it as _frozen_. Frozen peers are disregarded for all
+purposes except those discussed below - in particular no packets are sent to
+them except as described below, they are omitted from the peer lists sent to
+the client or in a Peer Response packet, and they are not considered when
+determining closest peers for establishing direct connections.
+
+A peer is considered to be active if we receive a group message or Rejoin
+packet from it, or a New Peer message for it.
+
+If a frozen peer is seen to be active, we remove its 'frozen' flag and send a
+Name group message. (We can hold off on sending this message until the next
+tox\_iterate, and only send one message if many frozen peers become active at
+once).
+
+If we receive a New Peer message for a peer, we update its DHT pubkey.
+
+If we receive a group message originating from an unknown peer, we drop the
+message but send a Peer Query packet back to the peer who directly sent us the
+message. (This is current behaviour; it's mentioned here because it's important
+and not currently mentioned in the spec.)
+
+If we receive a Rejoin packet from a peer we update its DHT pubkey, add a
+temporary groupchat connection for the peer, and, once the connection is
+online, send out a New Peer message announcing the peer, and a Name message.
+
+Whenever we make a new friend connection, we check if the public key is that 
+of any frozen peer. If so, we send it a Rejoin packet, add a temporary
+groupchat connection for it, and, once the connection is online, send the
+peer a Peer Query packet.
+
+We do the same with a peer when we are setting it as frozen if we have a
+friend connection to it.
+
+The temporary groupchat connections established in sending and handling Rejoin
+packets are not immediately operational (because group numbers are not known);
+rather, an Online packet is sent when we handle a Rejoin packet.
+
+When a connection is set as online as a result of an Online packet, we ping
+the group.
+
+When processing the reply to a Peer Query, we update the DHT pubkey of an
+existing peer if and only if it is frozen or has not had its DHT pubkey
+updated since it last stopped being frozen.
+
+When we receive a Title Response packet, we set the title if it has never been
+set or if at some point since it was last set, there were no unfrozen peers
+(except us).
+
+## Discussion
+### Overview
+The intention is to recover seamlessly from splits in the group, the most 
+common form of which is a single peer temporarily losing all connectivity.
+
+To see how this works, first note that groups (even before the changes
+discussed here) have the property that for a group to be connected in the
+sense that any peer will receive the messages of any other peer and have them
+in their peerlist, it is necessary and sufficient that there is a path of
+direct group connections between any two peers.
+
+Now suppose the group is split into two connected components, with each member
+of one component frozen according to the members of the other. Suppose there
+are two peers, one in each component, which are using the above protocol, and
+suppose they establish a friend connection. Then each will rejoin the other,
+forming a direct group connection. Hence the whole group will become connected
+(even if all other peers are using the unmodified protocol).
+
+The Peer Query packet sent on rejoining hastens this process.
+
+Peers who leave the group during a split will not be deleted by all peers
+after the merge - but they will be set as frozen due to ping timeouts, which
+is sufficient.
+
+### Titles
+If we have a split into components each containing multiple peers, and the
+title is changed in one component, then peers will continue to disagree on the
+title after the split. Short of a complicated voting system, this seems the
+only reasonable behaviour.
+
+### Implementation notes
+Although I've described the logic in terms of an 'frozen' flag, it might 
+actually make more sense in the implementation to have a separate list for 
+frozen peers.
+
+## Saving
+Saving could be implemented by simply saving all live groups with their group
+numbers and full peer info for all peers. On reload, all peers would be set as
+frozen.
+
+The client would need to support this by understanding that these groups exist
+on start-up (e.g. starting windows for them), and by not automatically killing
+groups on closing the client.
+
+## Limitations
+If a peer disconnects from the group for a period short enough that group
+timeouts do not occur, and a name change occurs during this period, then the
+name change will never be propagated.
+
+One way to deal with this would be a general mechanism for storing and
+requesting missed group messages. But this is considered out of scope of this
+PR.
+
+If a peer changes its DHT pubkey, the change might not be properly propagated
+under various circumstances - in particular, if connections do not go down
+long enough for the peer to become frozen.
+
+One way to deal with this would be to add a group message announcing the
+sending peer's current DHT pubkey, and treat it analogously to the Name
+message.

toxcore/group.h

@@ -49,8 +49,9 @@ typedef struct Message_Info {
 typedef struct Group_Peer {
     uint8_t     real_pk[CRYPTO_PUBLIC_KEY_SIZE];
     uint8_t     temp_pk[CRYPTO_PUBLIC_KEY_SIZE];
+    bool        temp_pk_updated;
 
-    uint64_t    last_recv;
+    uint64_t    last_active;
 
     Message_Info
     last_message_infos[MAX_LAST_MESSAGE_INFOS]; /* received messages, strictly decreasing in message_number */
@@ -79,11 +80,18 @@ typedef enum Groupchat_Close_Type {
     GROUPCHAT_CLOSE_ONLINE
 } Groupchat_Close_Type;
 
+/* Connection is to one of the closest DESIRED_CLOSE_CONNECTIONS peers */
+#define GROUPCHAT_CLOSE_REASON_CLOSEST     (1 << 0)
+
+/* Connection is to a peer we are introducing to the conference */
+#define GROUPCHAT_CLOSE_REASON_INTRODUCING (1 << 1)
+
+/* Connection is to a peer who is introducing us to the conference */
+#define GROUPCHAT_CLOSE_REASON_INTRODUCER  (1 << 2)
+
 typedef struct Groupchat_Close {
     uint8_t type; /* GROUPCHAT_CLOSE_* */
-    bool closest; /* connected to peer because it is one of our closest peers */
-    bool introducer; /* connected to peer because it introduced us to the group */
-    bool introduced; /* connected to peer because we introduced it to the group */
+    uint8_t reasons; /* bit field with flags GROUPCHAT_CLOSE_REASON_* */
     uint32_t number;
     uint16_t group_number;
 } Groupchat_Close;
@@ -101,9 +109,15 @@ typedef void group_on_delete_cb(void *object, uint32_t conference_number);
 typedef struct Group_c {
     uint8_t status;
 
+    bool need_send_name;
+    bool title_fresh;
+
     Group_Peer *group;
     uint32_t numpeers;
 
+    Group_Peer *frozen;
+    uint32_t numfrozen;
+
     /* TODO(zugz) rename close to something more accurate - "connected"? */
     Groupchat_Close close[MAX_GROUP_CONNECTIONS];
 
@@ -123,7 +137,7 @@ typedef struct Group_c {
 
     uint64_t last_sent_ping;
 
-    int number_joined; /* friendcon_id of person that invited us to the chat. (-1 means none) */
+    uint32_t num_introducer_connections;
 
     void *object;
 

toxcore/net_crypto.h

@@ -78,6 +78,7 @@
 #define PACKET_ID_ONLINE_PACKET 97
 #define PACKET_ID_DIRECT_CONFERENCE 98
 #define PACKET_ID_MESSAGE_CONFERENCE 99
+#define PACKET_ID_REJOIN_CONFERENCE 100
 #define PACKET_ID_LOSSY_CONFERENCE 199
 
 /*** Crypto connections. ***/