diff --git a/content/notes/21-SE-in-IA-2.md b/content/notes/21-SE-in-IA-2.md
index 54c198873..74167abe1 100644
--- a/content/notes/21-SE-in-IA-2.md
+++ b/content/notes/21-SE-in-IA-2.md
@@ -12,3 +12,56 @@ tags:
 - built without thinking about security
 - tests do not take security into account
 - without security evolution can become cumbersome
+
+more costly to fix bug in development than in design
+- the later we fix it the more costly it is
+
+from the start
+- hire right people
+- with hacking mindset
+	- so they alwasys think about security
+- invite customers to training and seminars
+	- testing could uncover security issues
+
+questions during feasability
+- ![](https://i.imgur.com/svo2sIb.png)
+- what are the implications for users if assets are lost
+	- some information is more high risk that other information
+
+during design
+- should be designed in iterative manner
+- threat modelling
+	- ![diagram](https://i.imgur.com/RM6lyC1.png)
+	- Data Flow Diagram (DFD)
+		- ![](https://i.imgur.com/iCEh2NU.png)
+		- high or low level
+		- want to be nimble
+		- dont want to go overboard
+		- ![level 1 dfd](https://i.imgur.com/rfyCUxH.png)
+		- 
+- potential threats
+- STRIDE
+	- ![list](https://i.imgur.com/ZBHK7v7.png)
+	- can identify parts system using DFD that are threatened by these things
+		- ![e.g](https://i.imgur.com/EwUSQzS.png)
+		- ![threat tree](https://i.imgur.com/RtqTzJ6.png)
+			- defines the potential dcisions that are performed by the attacker
+		- ![information to keep track of](https://i.imgur.com/k1vsDfu.png)
+		- risk = criticality * likligood of occurance
+			- should be consistent
+			- can also use DREAD
+				- ![list](https://i.imgur.com/VZYmjCU.png)
+
+example
+- ![](https://i.imgur.com/ZybxDCx.png)
+
+responsing to threats
+- do nothing
+- inform user
+- remove problem
+- fix problem
+
+
+
+
+