Track event sequences globally, not per channel.

Per-channel event sequences were a cute idea, but it made reasoning about event resumption much, much harder (case in point: recovering the order of events in a partially-ordered collection is quadratic, since it's basically graph sort). The minor overhead of a global sequence number is likely tolerable, and this simplifies both the API and the internals.

1 files changed, 44 insertions, 0 deletions

diff --git a/.sqlx/query-4715007e2395ad30433b7405a144db4901c302bbcd3e76da6c61742ac44345c9.json b/.sqlx/query-4715007e2395ad30433b7405a144db4901c302bbcd3e76da6c61742ac44345c9.json
new file mode 100644
index 0000000..494e1db
--- /dev/null
+++ b/.sqlx/query-4715007e2395ad30433b7405a144db4901c302bbcd3e76da6c61742ac44345c9.json
@@ -0,0 +1,44 @@
+{
+  "db_name": "SQLite",
+  "query": "\n\t\t\t\tinsert into message\n\t\t\t\t\t(id, channel, sender, sent_at, sent_sequence, body)\n\t\t\t\tvalues ($1, $2, $3, $4, $5, $6)\n\t\t\t\treturning\n\t\t\t\t\tid as \"id: message::Id\",\n\t\t\t\t\tsender as \"sender: login::Id\",\n                    sent_at as \"sent_at: DateTime\",\n                    sent_sequence as \"sent_sequence: Sequence\",\n\t\t\t\t\tbody\n\t\t\t",
+  "describe": {
+    "columns": [
+      {
+        "name": "id: message::Id",
+        "ordinal": 0,
+        "type_info": "Text"
+      },
+      {
+        "name": "sender: login::Id",
+        "ordinal": 1,
+        "type_info": "Text"
+      },
+      {
+        "name": "sent_at: DateTime",
+        "ordinal": 2,
+        "type_info": "Text"
+      },
+      {
+        "name": "sent_sequence: Sequence",
+        "ordinal": 3,
+        "type_info": "Integer"
+      },
+      {
+        "name": "body",
+        "ordinal": 4,
+        "type_info": "Text"
+      }
+    ],
+    "parameters": {
+      "Right": 6
+    },
+    "nullable": [
+      false,
+      false,
+      false,
+      false,
+      false
+    ]
+  },
+  "hash": "4715007e2395ad30433b7405a144db4901c302bbcd3e76da6c61742ac44345c9"
+}