diff --git a/documentation_file.txt b/documentation_file.txt
index c966e19..8e054ad 100644
--- a/documentation_file.txt
+++ b/documentation_file.txt
@@ -7,15 +7,19 @@ s4503678
 - etags using sha1sum
 - if no index.html: return 404
 - only GET support
-- persistence and timeouts and connection: close from both sides.
+- persistence and timeouts and 'connection: close' from both sides.
+- concurrency with pipelining
 
 implementation:
 
-implemented using sockets and select
+implemented in python using select. testing using 'unittest'.
+
 asyncserver.py contains the basic asynchronous server implementation;
 clients are stored in a list, and the event loop calls select() with a list of their sockets. Then, the clients are notified if there are any readable/writable/etc sockets that they own. The client connection class that is used can be specified when making the Server object.
 
 server.py builds on it to implement a http server, with HTTPClient implementing the basic HTTP server, and HTTPDirClient implementing the static file server.
 
+./run_tests.sh port requires the server to be active at the same port, it won't start it.
+
 difficulties:
 Testing took more time than desired, I wasn't sure if I was allowed to use the standard http library (or if it supports the neccesary tests without too much trouble), so I implemented some client functions.