diff --git a/hibernate-search/hsearch-elasticsearch-wikipedia/README.md b/hibernate-search/hsearch-elasticsearch-wikipedia/README.md
index f75243a..6014541 100644
--- a/hibernate-search/hsearch-elasticsearch-wikipedia/README.md
+++ b/hibernate-search/hsearch-elasticsearch-wikipedia/README.md
@@ -9,10 +9,24 @@ Create a database and user:
 ```
 # From a shell
 sudo -i -u postgres
-createuser -U postgres hibernate_demo # When prompted, use "hibernate_demo" as a password
+createuser -U postgres -P hibernate_demo # When prompted, use "hibernate_demo" as a password
 createdb -U postgres -O hibernate_demo hsearch_es_wikipedia
 ```

+You need to enable the `md5` authentication scheme for the local network connection. So, if you
+just installed PostgreSQL, you should make your `/var/lib/pgsql/data/pg_hba.conf` file look like:
+```
+local   all             all                                     peer
+host    all             all             127.0.0.1/32            md5
+host    all             all             ::1/128                 md5
+```
+(just change `ident` to `md5` in the existing lines)
+
+Don't forget to reload your PostgreSQL server after this change:
+```
+sudo systemctl reload postgresql
+```
+
 ## Data

 You can try a fully automated initialization using this command:
@@ -21,7 +35,10 @@ You can try a fully automated initialization using this command:
 ./src/init/init
 ```

-If this succeeds, you're all set, just run the project.
+The script will ask for a password: just use `hibernate_demo`.
+
+If this succeeds, you're all set, you can go to the next step.
+
 Otherwise, Wikipedia's dumps page structure probably changed, but you can proceed as explained below.

 Retrieve a dataset of current pages (one of those described as "All pages, current versions only.")
@@ -42,7 +59,6 @@ Initialize the database using the provided script:
 ```

 The script will ask for a password: just use `hibernate_demo`.
-Note that md5 mode authentication must be enabled in postgresql for this script to work.

 # Running the project