diff --git a/ex1/epsilon.cpp b/ex1/epsilon.cpp
index c6a209ccc42eafc737390c43157710dbecb8a967..440f5e2468e7bd4fe89cf207beae9046b4c329f7 100644
--- a/ex1/epsilon.cpp
+++ b/ex1/epsilon.cpp
@@ -1,3 +1,15 @@
+/*
+	Pascal: Looks good to me, albeit not necessarily precise.
+
+	Note that python's import is very different from preprocessor directives.
+	In python, there is a whole import machinery, while the #include directive
+	merely instructs the preprocessor to copy-paste a specific file.
+
+	Personally, I'm not a fan of 'using namespace xxx'. At the very least, you
+	pollute the global namespace and make it harder to keep track of where specific
+	functions and classes are actually coming from.
+*/
+
 // "include" is like the import in Python
 #include <iostream>
 // namespace allow us to group named entities into narrower scopes, that otherwise
@@ -5,7 +17,7 @@
 using namespace std;
 
 void machineEpsilon(float EPS) {
-	
+
 	float prev_epsilon = EPS;
 
 	while ((1+EPS) != 1){
@@ -20,4 +32,3 @@ void machineEpsilon(float EPS) {
 int main(){
 	machineEpsilon(0.5);
 }
-
diff --git a/ex1/main.cpp b/ex1/main.cpp
index c6cb9d29741fc061ffbd97ca5b0492c062bde00a..0d8eedaabdfde94cf6251b9fef80205902698227 100644
--- a/ex1/main.cpp
+++ b/ex1/main.cpp
@@ -1,3 +1,7 @@
+/*
+	Pascal: Looks good to me.
+*/
+
 # include <iostream>
 
 using namespace std;
diff --git a/ex1/simpson.cpp b/ex1/simpson.cpp
index 183612271855d047871a46c80ab06e56312251f2..e19a0c57d038eb3f6440a459df1249eeac78ba17 100644
--- a/ex1/simpson.cpp
+++ b/ex1/simpson.cpp
@@ -1,3 +1,12 @@
+/*
+	Pascal: Looks good to me.
+
+	You could still optimize the number of mathematical operations. For example,
+	per call to calculateSimpsonIntegral, you only need one multiplication with
+	binSize/6, as opposed to the N you are performing. You are also calling f
+	more often than necessary.
+*/
+
 //Simpson's integration method
 #include <iostream>
 #include <cmath>