Add getting started section for the extractor

TODO: rename all docs to place getting started at the beginning.

docs/08_getting_started.md

@@ -0,0 +1,89 @@
+# Getting Started
+This section provides you with the steps that are required to use the extractor.
+
+### Implement a Downloader
+
+The extractor needs to fetch websites and JSON files.
+Its entire network activity is going through the downloader you pass to it.  
+We have predefined some methods which need to be implemented.
+To get started, create a class extending
+[`Downloader`](https://github.com/TeamNewPipe/NewPipeExtractor/blob/master/extractor/src/main/java/org/schabi/newpipe/extractor/downloader/Downloader.java).
+We will call that class `DownloaderImpl` in this example.
+
+``` Java
+
+import org.schabi.newpipe.extractor.downloader.Downloader;
+
+public class DownloaderImpl extends Downloader {
+
+    public DownloaderImpl() {
+
+    }
+
+}
+
+```
+Your IDE will prompt you to implement at least one method:
+
+``` Java
+/**
+ * Do a request using the specified {@link Request} object.
+ *
+ * @return the result of the request
+ */
+public Response execute(@Nonnull Request request) throws IOException, ReCaptchaException {
+    // your code goes here
+}
+```
+
+You can find two example implementations of the Downloader below.
+Both use an old OkHttp version (version 3) to support Android 4. However, you do not need to use OkHttp.
+In case you want to use it and do not need to support old Android versions,
+we recommend using a newer OkHttp version, because the version used in the example will only receive security updates until December 2021.
+
+- [Downloader implementation for NewPipe Extractor tests](https://github.com/TeamNewPipe/NewPipeExtractor/blob/master/extractor/src/test/java/org/schabi/newpipe/downloader/DownloaderTestImpl.java)
+- [Downloader implementation of the NewPipe app](https://github.com/TeamNewPipe/NewPipe/blob/master/app/src/main/java/org/schabi/newpipe/DownloaderImpl.java)  
+Note: This implementation contains some additional methods which are not necessary for general use cases.
+
+
+Depending on your use case and your downloader implementation, we suggest you to make use of the singleton pattern.
+
+#### Browser Agent
+
+When making requests to websites, you should define a user agent for the request headers. We have build our tests on
+```
+Mozilla/5.0 (Windows NT 10.0; rv:78.0) Gecko/20100101 Firefox/78.0
+```
+For this reason, we strongly recommend using that agent or any other that is used in the [testimplementation in NewPipe Extractor](https://github.com/TeamNewPipe/NewPipeExtractor/blob/master/extractor/src/test/java/org/schabi/newpipe/downloader/DownloaderTestImpl.java).  
+However, a service might want to override the default user agent.
+This is sometimes necessary to get special versions of a website (e.g. mobile or legacy version).
+Please respect that when implementing the `execute(@Nonnull Request request)` method.
+
+## Initialize NewPipe
+
+``` Java
+NewPipe.init(
+    /* 
+      New instance of your downloader implementation.
+      If you implemented the singleton pattern, use DownlaoderImpl.getInstance() instead.
+    */
+    new DownloaderImpl()
+);
+```
+You can also set a default localization 
+``` Java
+NewPipe.init(
+    new DownloaderImpl(),
+    new Localization("EN", "US")
+);
+```
+and a default content country. However, not all services support those, yet.
+``` Java
+NewPipe.init(
+    new DownloaderImpl(),
+    new Localization("EN", "US"),
+    new ContentCountry("EN-US")
+);
+```
+
+