anonymization

Author: mircealungu

architecture.tex

@@ -63,9 +63,12 @@ \section{Architecture}
         \label{fig:sep}
     \end{figure}
 
-  As discussed in the introductory section, in previous work~\cite{vogel2017low} we introduced \tool, a drop-in Python library that allows developers to monitor their Flask-based Python web applications with minimal effort.
+  As discussed in the introductory section, in previous work \withheld we introduced \tool, a drop-in Python library that allows developers to monitor their Flask-based Python web applications with minimal effort.
 %
-  The \tool is implemented for Python 3.6 and is available on the Python Package Index repository\footnote{\url{https://pypi.python.org/pypi/flask-monitoring-dashboard/1.8}} from where it can be installed by running \install from the command line. 
+  The \tool is implemented for Python 3.6 and is available on the Python Package Index repository
+  % \footnote{\url{https://pypi.python.org/pypi/flask-monitoring-dashboard/1.8}} 
+  \footnote{\url{https://pypi.python.org/pypi/[\witheld]}} 
+  from where it can be installed by running \install from the command line. 
 %  
   The source code of the \tool is published under a permissive MIT license and is available on GitHub\footnote{\url{https://github.com/flask-dashboard}}.
 

evolution.tex

@@ -68,7 +68,7 @@ \subsection*{Evolving Grouped Performance}
 
 \begin{figure}[h!]
   \centering
-  \includegraphics[width=0.77\linewidth]{time_per_user_per_version}
+  \includegraphics[width=\linewidth]{time_per_user_per_version}
   \caption{This perspective shows that the evolution of response times for individual users (horizontal lines) across versions (the x-axis) for a given endpoint}
   \label{fig:tuv}
 \end{figure}

img/selecting_endpoints.png

@@ -72,7 +72,9 @@ \section{Case Study}
 \label{sec:api}
 \label{sec:case}
 
-  \zee\footnote{\url{https://github.com/zeeguu-ecosystem/}} is a platform and an ecosystem of applications for accelerating vocabulary acquisition in a foreign language \cite{Lungu16}. 
+  \zee
+  % \footnote{\url{https://github.com/zeeguu-ecosystem/}} 
+  is a platform and an ecosystem of applications for accelerating vocabulary acquisition in a foreign language \cite{lungu18-chi}. 
 %
   The architecture of the ecosystem has at its core an API and a series of ecosystem applications that together offer to a learner three main inter-dependent features:
 

img/selecting_endpoints_old.png

@@ -9,7 +9,7 @@
 
   The \tool is designed to take advantage of the fact that it runs in the same context as the web application it monitors, and that the monitored API already has a web presence. It thus simply makes available one extra endpoint (i.e. \code{/dashboard}) at the same address as the API, which serves the interactive application presented in the remainder of this paper. 
 
-  A dashboard can be attached to an already existing Flask application with a single line of code~\cite{vogel2017low} as the following snippet shows\footnote{We do not count the import statement that enables that line.}:
+  A dashboard can be attached to an already existing Flask application with a single line of code \citationwithheld as the following snippet shows\footnote{We do not count the import statement that enables that line.}:
 
 
   \begin{lstlisting}[style=custompython]
@@ -23,13 +23,11 @@
 
   During binding, the \tool will search for all endpoints defined in the target application and make them available in the dashboard in an interactive configuration panel (\Fref{fig:sep}). The panel presents all the automatically discovered endpoints and lets the user select the ones that should be monitored (the checkboxes in the last column). When the user enables the monitoring of an endpoint a function wrapper is added to the implementation of that endpoint (which is a function in Flask). The information is also saved in a database such that it is persistent across application restarts. 
 
-  \va{\Fref{fig:sep} needs to be updated --- it still has 2017 dates}
-
   % In order to monitor an endpoint, the \tool creates a function wrapper for the API function that corresponds to the endpoint. This way, the wrapper will be executed whenever that API call is made before the actual function is called. The wrapper contains the code that takes care of monitoring an endpoint. 
 
     \begin{figure}[h]
       \centering
-      \includegraphics[width=0.8\linewidth]{selecting_endpoints.png}
+      \includegraphics[width=\linewidth]{selecting_endpoints.png}
       \caption{Once connected to an API the Dashboard presents the endpoints that are available for monitoring}
       \label{fig:sep}
     \end{figure}

introductory.tex

@@ -4,7 +4,7 @@
 \section{Overhead of the \tool}
 \label{sec:overhead}
 
-To measure the performance overhead of the \tool, we have implemented an automated benchmarking system. It is open source and available online and can be tested by the reader. \va{@Mircea: Don't forget to add link} The benchmark downloads the latest version of the \zee API and installs it in a Docker container. Then it calls several selected endpoints for 500 times each, tracking the response times. The endpoints are called in three different configurations: 
+To measure the performance overhead of the \tool, we have implemented an automated benchmarking system. It is open source and available online and can be tested by the reader\footnote{\em [The link will be added after deanonymization]}. The benchmark downloads the latest version of the \zee API and installs it in a Docker container. Then it calls several selected endpoints for 500 times each, tracking the response times. The endpoints are called in three different configurations: 
 
 	\begin{enumerate}
 		\item With no dashboard installed.

loc1.tex

@@ -17,6 +17,16 @@ @inproceedings{vogel2017low
 }
 
 
+@inproceedings{lungu18-chi,
+  author={Mircea F. Lungu and Dan Chirtoaca and Martin Avagyan and Luc van den Brand},
+  title = {{As We May Study: Towards the Web as a Personalized Language Textbook}},
+  date = {2017},
+  booktitle ={{Proceedings of CHI'18: ACM Conference on Human Factors in Computing Systems}},
+  pages = {(to appear)},
+  publisher ={ACM},
+}
+
+
 # Related Work Papers
 
 

overhead.tex

@@ -1,7 +1,7 @@
 %!TEX root=paper.tex
 
 
-
+\vspace{1cm}
   \section{Performance Triangulation With Regression Monitoring}
   \label{sec:testing}
 
@@ -31,7 +31,7 @@
 
       \begin{figure}[h!]
         \centering
-        \includegraphics[width=0.65\columnwidth]{travis_builds-bw}
+        \includegraphics[width=0.85\columnwidth]{travis_builds-bw}
         \caption{Response times in 5 consequent Travis builds}
         \label{fig:builds}
       \end{figure}

paper.bib

@@ -30,6 +30,9 @@
 \newcommand{\activeUserCount}{two hundred\xspace}
 \newcommand{\code}[1]{\texttt{#1}\xspace}
 \newcommand{\perspective}[1]{{\bf {\small {\texttt{#1}}\xspace}}}
+\newcommand{\withheld}{withheld due to double blind}
+\newcommand{\citationwithheld}{{\em [citation withheld due to double blind]}}
+
 
 \newcommand{\lesson}[1]{{\bf {\small {\texttt{Lesson: }}\xspace}}: #1}
 

testing.tex

@@ -12,7 +12,7 @@ \subsection{Utilization}
 
     \begin{figure}[h!]
       \centering
-      \includegraphics[width=0.8\linewidth]{utilization-table}
+      \includegraphics[width=.97\linewidth]{utilization-table}
       \caption{The API Overview presents synthetic endpoint utilization information}
       \label{fig:basicest}
     \end{figure}
@@ -38,7 +38,7 @@ \subsection{Utilization}
   \end{itemize}
 
 
-\niceseparator
+% \niceseparator
 
 
   For a more detailed view of utilization evolution, \Fref{fig:aeu} shows the \perspective{Daily API Utilization} perspective on API utilization that \tool provides: a stacked bar chart of the number of hits to various endpoints grouped by day\footnote{Endpoint colors are the same in different views}. \Fref{fig:aeu} in particular shows a peak utilization, on June 28, 2017 when the API had more than 12.000 hits\footnote{Turns out that the high school students that were using the system for their French-learning class had a deadline on that date}.