Prepare new release v0.14.0.

CHANGELOG.md

@@ -1,3 +1,17 @@
+## v0.14.0
+
+* Allow truly random UUIDv4 record IDs by relying on AES encrypted `_rowid_`s
+  for cursors. UUIDv7, while great, has the problem of leaking creation-time.
+  * Note: the encryption key for cursors is tied to the instance lifetime, i.e.
+    they cannot be used across instance restarts (at least for now).
+* Move user ids to UUIDv4 by default to avoid leaking user creation-time.
+  * A bundled schema migration will update the PK to allow any UUID type, this is
+    mostly to allow for existing users with UUIDv7 ids to continue to exist.
+  * We expect this change to be transparent for most users but may break you,
+    if you're relying on user ids being of the UUIDv7. Sorry, we felt this was an
+    important change and wanted to rip off the band-aid. If you're having issues
+    and are unsure on how to address them, please reach out and we'll help.
+
 ## v0.13.3
 
 * Improve RecordAPI settings sheet with tabs, help text, and curl examples.

docs/src/content/docs/documentation/APIs/record_apis.mdx

@@ -13,8 +13,8 @@ TrailBase's restful CRUD _Record APIs_ allow you to access your `TABLE`s and
 - Tables and views need to be `STRICT`ly[^1] typed to guarantee type-safety all the
   way from your records, via JSON schema, to your client-side language bindings [^2].
 - They need to have a sequential primary key column to allow for stable sorting
-  and thus efficient cursor-based pagination. Either an explicit `INTEGER` or
-  UUIDv7 `PRIMARY KEY` will do, including `FOREIGN KEY` columns.
+  and thus efficient cursor-based pagination. Either an explicit `INTEGER`,
+  UUIDv4 or UUIDv7 `PRIMARY KEY` will do, including `FOREIGN KEY` columns.
 
 ## Configuration
 

docs/src/content/docs/getting-started/first-cli-app.mdx

@@ -40,7 +40,7 @@ CREATE TABLE IF NOT EXISTS movies (
 ```
 
 Note that the only schema requirement for exposing an API is: `STRICT` typing
-and an integer (or UUIDv7) primary key column.
+and an integer, UUIDv4 or UUIDv7 primary key column.
 
 The main benefit of relying on TrailBase to apply the above schema as migrations
 over manually applying the schema yourself, is to: