Edits based on Ryan's feedback

Author: Lauren

v19.2/changefeed-for.md

@@ -21,8 +21,9 @@ Changefeeds can only be created by superusers, i.e., [members of the `admin` rol
 ## Considerations
 
 - Because core changefeeds return results differently than other SQL statements, they require a dedicated database connection with specific settings around result buffering. In normal operation, CockroachDB improves performance by buffering results server-side before returning them to a client; however, result buffering is automatically turned off for core changefeeds. Core changefeeds also have different cancelation behavior than other queries: they can only be canceled by closing the underlying connection or issuing a [`CANCEL QUERY`](cancel-query.html) statement on a separate connection. Combined, these attributes of changefeeds mean that applications should explicitly create dedicated connections to consume changefeed data, instead of using a connection pool as most client drivers do by default.
-    This cancelation behavior also extends to client driver usage; in particular, when a client driver calls `Rows.Close()` after encountering errors for a stream of rows. The pgwire protocol requires that the rows be consumed before the connection is again usable, but in the case of a core changefeed, the rows are never consumed. It is therefore critical that you close the connection, otherwise the application will be blocked forever on `Rows.Close()`.
-    
+
+    This cancelation behavior (i.e., close the underlying connection to cancel the changefeed) also extends to client driver usage; in particular, when a client driver calls `Rows.Close()` after encountering errors for a stream of rows. The pgwire protocol requires that the rows be consumed before the connection is again usable, but in the case of a core changefeed, the rows are never consumed. It is therefore critical that you close the connection, otherwise the application will be blocked forever on `Rows.Close()`.
+
 - In most cases, each version of a row will be emitted once. However, some infrequent conditions (e.g., node failures, network partitions) will cause them to be repeated. This gives our changefeeds an at-least-once delivery guarantee. For more information, see [Change Data Capture - Ordering Guarantees](change-data-capture.html#ordering-guarantees).
 ## Synopsis
 

v20.1/changefeed-for.md

@@ -21,7 +21,8 @@ Changefeeds can only be created by superusers, i.e., [members of the `admin` rol
 ## Considerations
 
 - Because core changefeeds return results differently than other SQL statements, they require a dedicated database connection with specific settings around result buffering. In normal operation, CockroachDB improves performance by buffering results server-side before returning them to a client; however, result buffering is automatically turned off for core changefeeds. Core changefeeds also have different cancelation behavior than other queries: they can only be canceled by closing the underlying connection or issuing a [`CANCEL QUERY`](cancel-query.html) statement on a separate connection. Combined, these attributes of changefeeds mean that applications should explicitly create dedicated connections to consume changefeed data, instead of using a connection pool as most client drivers do by default.
-    This cancelation behavior also extends to client driver usage; in particular, when a client driver calls `Rows.Close()` after encountering errors for a stream of rows. The pgwire protocol requires that the rows be consumed before the connection is again usable, but in the case of a core changefeed, the rows are never consumed. It is therefore critical that you close the connection, otherwise the application will be blocked forever on `Rows.Close()`.
+
+    This cancelation behavior (i.e., close the underlying connection to cancel the changefeed) also extends to client driver usage; in particular, when a client driver calls `Rows.Close()` after encountering errors for a stream of rows. The pgwire protocol requires that the rows be consumed before the connection is again usable, but in the case of a core changefeed, the rows are never consumed. It is therefore critical that you close the connection, otherwise the application will be blocked forever on `Rows.Close()`.
 
 - In most cases, each version of a row will be emitted once. However, some infrequent conditions (e.g., node failures, network partitions) will cause them to be repeated. This gives our changefeeds an at-least-once delivery guarantee. For more information, see [Change Data Capture - Ordering Guarantees](change-data-capture.html#ordering-guarantees).
 

v20.2/changefeed-for.md

@@ -20,6 +20,14 @@ For more information, see [Change Data Capture](change-data-capture.html).
 
 Changefeeds can only be created by superusers, i.e., [members of the `admin` role](authorization.html#create-and-manage-roles). The admin role exists by default with `root` as the member.
 
+## Considerations
+
+- Because core changefeeds return results differently than other SQL statements, they require a dedicated database connection with specific settings around result buffering. In normal operation, CockroachDB improves performance by buffering results server-side before returning them to a client; however, result buffering is automatically turned off for core changefeeds. Core changefeeds also have different cancelation behavior than other queries: they can only be canceled by closing the underlying connection or issuing a [`CANCEL QUERY`](cancel-query.html) statement on a separate connection. Combined, these attributes of changefeeds mean that applications should explicitly create dedicated connections to consume changefeed data, instead of using a connection pool as most client drivers do by default.
+
+    This cancelation behavior (i.e., close the underlying connection to cancel the changefeed) also extends to client driver usage; in particular, when a client driver calls `Rows.Close()` after encountering errors for a stream of rows. The pgwire protocol requires that the rows be consumed before the connection is again usable, but in the case of a core changefeed, the rows are never consumed. It is therefore critical that you close the connection, otherwise the application will be blocked forever on `Rows.Close()`.
+
+- In most cases, each version of a row will be emitted once. However, some infrequent conditions (e.g., node failures, network partitions) will cause them to be repeated. This gives our changefeeds an at-least-once delivery guarantee. For more information, see [Change Data Capture - Ordering Guarantees](change-data-capture.html#ordering-guarantees).
+
 ## Synopsis
 
 ~~~