fixes:

1. DOCKER_HOST is a variable, treat it like one
2. Adding htop, because it's especially linuxy and useful for a demo
3. Fixed readme grammar/language

Dockerfile

@@ -1,6 +1,6 @@
 FROM ubuntu:14.04
 RUN apt-get update -q
-RUN apt-get install -qy npm ssh
+RUN apt-get install -qy npm ssh htop
 RUN command -v node >/dev/null 2>&1 || { ln -s /usr/bin/nodejs /usr/bin/node; }
 RUN npm install manta -g
 ADD ./bin /usr/local/sbin

README.md

@@ -8,7 +8,7 @@ It's likely that all benchmarking tools suffer fundamental flaws, this is no exc
 
 > If you've spent less than a week studying a benchmark result, it's probably wrong.
 
-If that caught your attention, read the Summary near the end of [Brendan's blog post](http://www.brendangregg.com/ActiveBenchmarking/bonnie++.html#summary), then read more if that caught your attention. The quotes come from that post, and Brendan explains them in much more detail there.
+If that caught your attention, read the Summary near the end of [Brendan's blog post](http://www.brendangregg.com/ActiveBenchmarking/bonnie++.html#summary), then read the reast. The quotes come from that post, and Brendan explains them in much more detail there.
 
 The best way to benchmark a system is to run your app on it, since no other code will behave quite the same way.
 

bin/simple-container-benchmarks

@@ -22,7 +22,7 @@ echo '------------------------------' >> ~/simple-container-benchmarks.txt
 echo 'Performance benchmarks' >> ~/simple-container-benchmarks.txt
 echo '------------------------------' >> ~/simple-container-benchmarks.txt
 echo 'hostname: '$(hostname) >> ~/simple-container-benchmarks.txt
-echo "dockerhost: DOCKER_HOST" >> ~/simple-container-benchmarks.txt
+echo "dockerhost: $DOCKER_HOST" >> ~/simple-container-benchmarks.txt
 echo 'date: '$(date) >> ~/simple-container-benchmarks.txt