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, 38 insertions, 0 deletions

diff --git a/.sqlx/query-9386cdaa2cb41f5a7e19d2fc8c187294a4661c18c2d820f4379dfd82138a8f77.json b/.sqlx/query-9386cdaa2cb41f5a7e19d2fc8c187294a4661c18c2d820f4379dfd82138a8f77.json
new file mode 100644
index 0000000..e9c3967
--- /dev/null
+++ b/.sqlx/query-9386cdaa2cb41f5a7e19d2fc8c187294a4661c18c2d820f4379dfd82138a8f77.json
@@ -0,0 +1,38 @@
+{
+  "db_name": "SQLite",
+  "query": "\n                select\n                    id as \"id: Id\",\n                    name,\n                    created_at as \"created_at: DateTime\",\n                    created_sequence as \"created_sequence: Sequence\"\n                from channel\n                where coalesce(created_sequence > $1, true)\n            ",
+  "describe": {
+    "columns": [
+      {
+        "name": "id: Id",
+        "ordinal": 0,
+        "type_info": "Text"
+      },
+      {
+        "name": "name",
+        "ordinal": 1,
+        "type_info": "Text"
+      },
+      {
+        "name": "created_at: DateTime",
+        "ordinal": 2,
+        "type_info": "Text"
+      },
+      {
+        "name": "created_sequence: Sequence",
+        "ordinal": 3,
+        "type_info": "Integer"
+      }
+    ],
+    "parameters": {
+      "Right": 1
+    },
+    "nullable": [
+      false,
+      false,
+      false,
+      false
+    ]
+  },
+  "hash": "9386cdaa2cb41f5a7e19d2fc8c187294a4661c18c2d820f4379dfd82138a8f77"
+}