README.md in telegram-bot-0.9.0.alpha2 vs README.md in telegram-bot-0.9.0

@@ -15,11 +15,12 @@
   [httpclient](https://github.com/nahi/httpclient) under the hood).
 - Controller with message parser. Allows to write separate methods for each command.
 - Middleware and routes helpers for production env.
 - Poller with automatic source-reloader for development env.
 - Rake tasks to update webhook urls.
-- Async requests for Telegram and/or Botan API. Let the queue adapter handle errors!
+- __[Async mode](#async-mode)__ for Telegram and/or Botan API.
+  Let the queue adapter handle network errors!
 
 Here is sample [telegram_bot_app](https://github.com/telegram-bot-rb/telegram_bot_app)
 with session, keyboards and inline queries.
 Run it on your local machine in 1 minute!
 
@@ -130,10 +131,11 @@
   # `on_` to avoid conflicts.
   def start(data = nil, *)
     # do_smth_with(data)
 
     # There are `chat` & `from` shortcut methods.
+    # For callback queries `chat` if taken from `message` when it's available.
     response = from ? "Hello #{from['username']}!" : 'Hi there!'
     # There is `respond_with` helper to set `chat_id` from received message:
     respond_with :message, text: response
     # `reply_with` also sets `reply_to_message_id`:
     reply_with :photo, photo: File.open('party.jpg')
@@ -244,11 +246,11 @@
   context_to_action!
   # It'll use context value as action name for all contexts which miss handlers.
 end
 ```
 
-You can use `CallbackQueyContext` in the similar way to split `#callback_query` into
+You can use `CallbackQueryContext` in the similar way to split `#callback_query` into
 several specific methods. It doesn't require session support, and takes context from
 data. If data has a prefix with colon like this `my_ctx:smth...` it'll call
 `my_ctx_callback_query('smth...')` when there is such action method. Otherwise
 it'll call `callback_query('my_ctx:smth...')` as usual.
 
@@ -332,10 +334,27 @@
 
 There are also some helpers for controller tests.
 Check out `telegram/bot/updates_controller/rspec_helpers` and
 `telegram/bot/updates_controller/testing`.
 
+Built-in RSpec matchers will help you to write tests fast:
+
+```ruby
+include Telegram::Bot::RSpec::ClientMatchers # no need if you already use controller herlpers
+
+expect(&process_update).to send_telegram_message(bot, /msg regexp/, some: :option)
+expect(&process_update).
+  to make_telegram_request(bot, :sendMessage, hash_including(text: 'msg text'))
+
+# controller specs are even simplier:
+describe '#start' do
+  subject { -> { dispatch_message '/start' } }
+  it { should respond_with_message(/Hello/) }
+end
+# See sample app for more examples.
+```
+
 ### Deploying
 
 Use `rake telegram:bot:set_webhook` to update webhook url for all configured bots.
 Certificate can be specified with `CERT=path/to/cert`.
 
@@ -382,13 +401,30 @@
 
 - Process updates very fast, without waiting for telegram and botan responses.
 - Handle and retry network and other errors with queue adapter.
 - ???
 
+Instead of performing request instantly client serializes it, pushes to queue,
+and immediately return control back. The job is then fetched with a worker
+and real API request is performed. And this all is absolutely transparent for the app.
+
 To enable this mode add `async: true` to bot's and botan's config.
 For more information and custom configuration check out
 [docs](http://www.rubydoc.info/github/telegram-bot-rb/telegram-bot/master/Telegram/Bot/Async) or
 [source](https://github.com/telegram-bot-rb/telegram-bot/blob/master/lib/telegram/bot/async.rb).
+
+If you want async mode, but don't want to setup queue, know that Rails 5 are shipped
+with Async adapter by default, and there is
+[Sucker Punch](https://github.com/brandonhilkert/sucker_punch) for Rails 4.
+
+Be aware of some limitations:
+
+- Client will not return API response.
+- Sending files is not available in async mode [now],
+  because them can not be serialized.
+
+To disable async mode for the block of code use `bot.async(false) { bot.send_photo }`.
+Yes, it's threadsafe too.
 
 ## Development
 
 After checking out the repo, run `bin/setup` to install dependencies.
 Then, run `rake spec` to run the tests.