diff --git a/README.md b/README.md
index e4ef0f51..144be6c7 100644
--- a/README.md
+++ b/README.md
@@ -43,7 +43,7 @@ Java-version of GraphChi: https://github.com/GraphChi/graphchi-java
 
 ### Publication
 
-GraphChi is part of the OSDI'12 proceedings. PDF of the paper can be downloaded here: http://select.cs.cmu.edu/publications/paperdir/osdi2012-kyrola-blelloch-guestrin.pdf
+GraphChi is part of the OSDI'12 proceedings. PDF of the paper can be downloaded here: https://www.usenix.org/system/files/conference/osdi12/osdi12-final-126.pdf
 
 Slides (OSDI talk): http://www.cs.cmu.edu/~akyrola/files/osditalk-graphchi.pptx
 
@@ -55,7 +55,6 @@ http://bickson.blogspot.com/2012/12/collaborative-filtering-with-graphchi.html
 ## Features
 
 - Vertex-centric computation model (similar to GraphLab, Pregel or Giraph)
-  ** Wrapper for GraphLab 2.1 API (Gather-Apply-Scatter model)
 - Asynchronous, parallel execution, with (optional) deterministic scheduling (see semantics section at [Creating-GraphChi-Applications](https://github.com/GraphChi/graphchi-cpp/wiki/Creating-GraphChi-Applications) )
 
 - Can run graphs with billions of edges, with linear scalability, on a standard consumer grade machine
@@ -70,7 +69,7 @@ http://bickson.blogspot.com/2012/12/collaborative-filtering-with-graphchi.html
 
 ## Getting Started
 
-Best way to get started is to start from the ExampleApps page.
+Best way to get started is to start from the https://github.com/GraphChi/graphchi-cpp/wiki/Example-Apps page.
 Prior to that, you need to download the source code (no configuration
 or installation is required).
 
@@ -131,10 +130,6 @@ While distributed clusters can solve the same problems faster than GraphChi on a
   <tr><td>5</td><td> S. Suri and S. Vassilvitskii. Counting triangles and the curse of the last reducer. In Proceedings of the 20th international conference on World wide web, pages 607–614. ACM, 2011. </td></tr>
 </table>
 
-### Comparison to Giraph 
-
-Apache [Giraph](http://giraph.apache.org/) is an open-source implementation of the Pregel graph engine, built on top of Hadoop. Based on a recent talk by the main developer of Giraph ( http://www.youtube.com/watch?v=b5Qmz4zPj-M ), Giraph running with 20 workers can run five iterations of !PageRank  on a graph with 5 billion edges in approx. 75 minutes. Estimated from our results above, GraphChi can execute similar task in roughly the same time on just one machine. The structure of the input graph affects the runtime, so a direct comparison is not possible without using the same input.  
-
 ### Preprocessing times
 
 <table>
diff --git a/README.txt b/README.txt
index 85be0d30..de7bf04b 100644
--- a/README.txt
+++ b/README.txt
@@ -4,7 +4,7 @@ GraphChi is a powerful graph computation engine that can process very
 large graphs from the disk.
 
 For documentation and background, please visit: 
-       https://code.google.com/p/graphchi/
+       https://github.com/GraphChi/graphchi-cpp
 
 
 INSTALLATION
@@ -15,7 +15,7 @@ GraphChi does not require installation, it is headers-only. Simply type
 GETTING STARTED
 
 Start from the example applications. See instructions here:
-      https://code.google.com/p/graphchi/wiki/ExampleApps
+      https://github.com/GraphChi/graphchi-cpp/wiki/Example-Apps
 
 ... or simply type "make apps".
 
diff --git a/src/util/labelanalysis.hpp b/src/util/labelanalysis.hpp
index b87a4fd6..8982d94d 100644
--- a/src/util/labelanalysis.hpp
+++ b/src/util/labelanalysis.hpp
@@ -166,7 +166,7 @@ void analyze_labels(std::string basefilename, int printtop = 20) {
     std::string outname = basefilename + ".components";
     std::ofstream resf;
     resf.open(outname.c_str());
-    if (resf == NULL) {
+    if (resf.fail()) {
         logstream(LOG_ERROR) << "Could not write label outputfile : " << outname << std::endl;
         return;
     }