diff --git a/README.md b/README.md
index cfef2ea..285deed 100644
--- a/README.md
+++ b/README.md
@@ -10,12 +10,6 @@ Grrrrrrrrrrls is a project in progress for the [Computer Grrrls](https://hmkv.de
 
     $ pip3 install nltk
 
-# Txt documents 
-
-The search machine is using the index.json file to process results. The function 'create_index' can be called to generate this file. It uses a set of plain text files to index each word and its corresponding TFIDF value. The plain text files are not included in this repo, i don't think i can publish them like that. 
-
-If you want to work with another set of documents, make a 'txt/' folder, add a few txt files in it, and remove the index.json file (or rename it if you want to keep it with you).
-
 # Start
 
 Start the flask/python local server ...
@@ -26,6 +20,29 @@ Browse to your localhost on port 5000 ...
 
     > 127.0.0.1:5000
 
+
+# Txt documents 
+
+The search machine is using the index.json file to process results. The function 'create_index' can be called to generate this file. It uses a set of plain text files to index each word and its corresponding TFIDF value. The plain text files are not included in this repo, i don't think i can publish them like that. 
+
+
+# Changing txt documents
+
+If you want to work with another set of texts, make a 'txt/' folder, add a few txt files in it, and remove the index.json file (or rename it if you want to keep it with you).
+
+To generate a new index.json file: 
+
+Remove the index.json file
+
+    $ rm index.json
+
+Stop and start the python server...
+
+    ctrl + c
+
+    $ python3 start.py
+
+
 ## Notes
 
 This Grrrrrrrrrrls search machine cannot handle too much at once: it can only work with one word.