diff --git a/Working-with-Updates.md b/Working-with-Updates.md
index 929c80d..d7880a0 100644
--- a/Working-with-Updates.md
+++ b/Working-with-Updates.md
@@ -41,7 +41,25 @@ We only ask you one thing: don't keep this running for too long, or your contact
 
 ## Spawning no worker at all
 
-**WIP**
+All the workers do is loop forever and poll updates from a queue that is filled from the `ReadThread`, responsible for reading every item off the network. If you only need a worker and the `MainThread` would be doing no other job, this is the preferred way. You can easily do the same as the workers like so:
+
+```python
+while True:
+    update = client.updates.poll()
+    if not update:
+        continue
+
+    print('I received', update)
+```
+
+Note that `poll` accepts a `timeout=` parameter, and it will return `None` if other thread got the update before you could or if the timeout expired, so it's important to check `if not update`.
+
+This can coexist with the rest of `N` workers, or you can set it to `0` additional workers:
+```python
+client = TelegramClient('session', api_id, api_hash, update_workers=0)
+```
+
+You **must** set it to `0` (or other number), as it defaults to `None` and there is a different. `None` workers means updates won't be processed *at all*, so you must set it to some value (0 or greater) if you want `client.updates.poll()` to work.
 
 ## Using the main thread instead the `ReadThread`
 
@@ -65,7 +83,7 @@ As a complete example:
 def callback(update):
     print('I received', update)
 
-client = TelegramClient('anon', api_id, api_hash,
+client = TelegramClient('session', api_id, api_hash,
                         update_workers=1, spawn_read_thread=False)
 
 client.connect()