Clean up examples and tutorials as well as minor doc tweaks.

Also add a reconnect loop to collab clicker.

examples/data-cli-tutorial/data/README.md

@@ -0,0 +1,35 @@
+# About the Dataset.
+
+This dataset was originally compiled from [IMDB
+data](https://developer.imdb.com/non-commercial-datasets/) and is subject to
+their terms of use and any applicable legal restrictions..
+
+The compilation is provided by
+[kaggle](https://www.kaggle.com/datasets/inductiveanks/top-1000-imdb-movies-dataset/data)
+and can be downloaded via:
+
+```bash
+curl -o archive.zip https://www.kaggle.com/api/v1/datasets/download/inductiveanks/top-1000-imdb-movies-dataset
+```
+
+## Schema
+
+```sql
+CREATE TABLE movies (
+  rank         INTEGER PRIMARY KEY,
+  name         TEXT NOT NULL,
+
+  -- Year cannot be INTEGER, since some are like "I 2016".
+  year         ANY NOT NULL,
+  watch_time   INTEGER NOT NULL, -- in minutes
+  rating       REAL NOT NULL,
+
+  -- Ideally nullable integer, however sqlite assumes empty to be text.
+  metascore    ANY,
+
+  -- Ideally nullable real, however sqlite assumes empty to be text.
+  gross        ANY,
+  votes        TEXT NOT NULL,
+  description  TEXT NOT NULL
+) STRICT;
+```