Make wicket encode its fortunes response in UTF-8 (#2664)

It appeared to be using the system default charset instead, which was
not necessarily UTF-8.

I removed the initialCapacity argument for the ArrayList because there's
no good reason to set it to 10,000.  Since the test requirements state
that the list shouldn't be sized using foreknowledge of the row count,
we might as well trust ArrayList's defaults.

frameworks/Java/wicket/src/main/java/hellowicket/fortune/FortunePage.java

@@ -23,10 +23,10 @@ import org.apache.wicket.request.http.WebResponse;
  */
 public class FortunePage extends WebPage {
 	private static final long serialVersionUID = 1L;
-	private static final String TEXT_HTML = "text/html";
+	private static final String TEXT_HTML = "text/html;charset=utf-8";
 
 	public FortunePage() throws Exception {
-		List<Fortune> fortunes = new ArrayList<>(10000);
+		List<Fortune> fortunes = new ArrayList<>();
 
 		DataSource dataSource = WicketApplication.get().getDataSource();
 		try ( //