diff --git a/README.md b/README.md
index 934f202..cb6bba6 100644
--- a/README.md
+++ b/README.md
@@ -77,6 +77,11 @@ x=gf.fFort(SHARED_LIB_NAME,MOD_FILE_NAME)
 
 ``x`` now contains all variables, parameters and procedures from the module (tab completable). 
 
+> **_NOTE:_** The mod data is cached to speed up re-reading the data. To control this pass cache_folder to fFort.
+A value of False disables caching, a string sets the folder location, while leaving the argument as None defaults to appdirs ``user_cache_dir``.
+
+
+
 ### Functions
 ````python
 y = x.func_name(a,b,c)
diff --git a/tests/README.md b/tests/README.md
index 68e277c..36174bc 100644
--- a/tests/README.md
+++ b/tests/README.md
@@ -1,6 +1,6 @@
 # How many tests do we actually?
 
-The main difficulty in writing gfort2py is handling ths head number of combinations of options that are available in Fortran code. 
+The main difficulty in writing gfort2py is handling the huge number of combinations of options that are available in Fortran. 
 If we consider a simple integer: How many different ways can we declare an integer?
 
 First we must consider where the integer is declared:
@@ -24,7 +24,8 @@ Is it an array? If so:
 - Explicit but runtime sized (dimension(n) where n is another integer)
 - Assumed size (dimension(*))
 - Assumed shape (dimension(:))
-- Allocatable
+- allocatable
+- pointer/target
 
 Then of course there are the combinations of the previous options. As well, for arrays its helpful to test multiple dimensions to ensure the
 ordering is correct.
@@ -154,7 +155,7 @@ test similar things should have similar names, i.e all thing testing explicit ar
 
 # Skipping tests
 
-For things that don't work yet you can use the python decorator ``@pytest.mark.skip`` to skip the test. NOthing needs to be done on the Fortran side for tests that don't work yet.
+For things that don't work yet you can use the python decorator ``@pytest.mark.skip`` to skip the test. Nothing needs to be done on the Fortran side for tests that don't work yet.
 
 # Bug reports