diff --git a/mPyPl/core.py b/mPyPl/core.py
index ed9f668..bf5b40e 100644
--- a/mPyPl/core.py
+++ b/mPyPl/core.py
@@ -29,11 +29,15 @@ def apply(datastream, src_field, dst_field, func,eval_strategy=None):
     """
     Applies a function to the specified field of the stream and stores the result in the specified field. Sample usage:
     `[1,2,3] | as_field('f1') | apply('f1','f2',lambda x: x*x) | select_field('f2') | as_list`
+    If `dst_field` is `None`, function is just executed on the source field(s), and result is not stored.
+    This is useful when there are side effects.
     """
     def applier(x):
-        x[dst_field] =  (lambda : __fnapply(x,src_field,func)) if lazy_strategy(eval_strategy) else __fnapply(x,src_field,func)
-        if eval_strategy:
-            x.set_eval_strategy(dst_field,eval_strategy)
+        r = (lambda : __fnapply(x,src_field,func)) if lazy_strategy(eval_strategy) else __fnapply(x,src_field,func)
+        if dst_field is not None and dst_field!='':
+            x[dst_field]=r
+            if eval_strategy:
+                x.set_eval_strategy(dst_field,eval_strategy)
         return x
     return datastream | select(applier)
 
diff --git a/mPyPl/sink.py b/mPyPl/sink.py
index 90b1177..825aa13 100644
--- a/mPyPl/sink.py
+++ b/mPyPl/sink.py
@@ -3,15 +3,18 @@
 
 # Different sinks to consume mdict streams
 
+from pipe import *
 import csv
 import json
 
+@Pipe
 def write_csv(l,filename):
-    with open(filename,'wb') as f:
+    with open(filename,'a') as f:
         w = csv.writer(f)
         for x in l:
             w.writerow(x)
 
+@Pipe
 def write_json(l,filename):
     with open(filename,'w') as f:
         f.write(json.dumps(l))