deletions | additions       diff --git a/by-example.tex b/by-example.tex  index 544894c..dc89e86 100644  --- a/by-example.tex  +++ b/by-example.tex 
...
In this section, we illustrate the key condition, components,  dimensions, and environmental factors that support serendipity, using  historical examples. The structure of this section follows and  updates an earlier survey from \citeA{pease2013discussion}, and  prepares the way for our model. \citeA{pease2013discussion}.  \subsubsection*{Key condition for serendipity}  diff --git a/conclusion.tex b/conclusion.tex  index a168d12..5643d4a 100644  --- a/conclusion.tex  +++ b/conclusion.tex 
...
\section{Conclusion} \label{sec:conclusion}  % What answers have we offered?  The ideas presented in this article present outline  several interesting possible  directions for implementation, but in any case  considerable concrete work remains to be done in order to realise the our  model in code. We examined a number of prior computational systems that match some aspects of our model, but even these hand-picked examples pale in comparison to the examples of serendipitous discovery and invention  from human history. It would seem that a fully-automated system that can realistically be said to behave in a serendipitous manner has not yet been built. % Further questions  The Nevertheless, the  theoretical work in this paper shows that it is indeed possible to plan -- and program -- for serendipity. diff --git a/future-work-conclusion.tex b/future-work-conclusion.tex  index 1f71120..f71af05 100644  --- a/future-work-conclusion.tex  +++ b/future-work-conclusion.tex 
...
this sort of system is our next planned step. Representing, and  learning, design patterns in programmatic terms is a ``stretch goal''  -- but patterns like the one above will offer useful heuristic  guidelines for human programmers right away. away, and give an idea  of our long-term plans.  % Is ``having a stretch goal'' an example of a serendipity pattern? I think so!  diff --git a/future-work-intro.tex b/future-work-intro.tex  index 0e0442f..4622d32 100644  --- a/future-work-intro.tex  +++ b/future-work-intro.tex 
...
addition describing the solution (the `how').'' Regarding the  criteria that pattern writers seek to address: ``The most appropriate  solution to a problem in a context is the one that best resolves the  highest priority forces as determined by the particular context.'' %% Their article describes a number of criteria%%  relevant to writing %%  good design patterns, e.g. \emph{Clear target%%  audience}, %%  \emph{Visible forces}, and \emph{Relationship to other%%  patterns}. A good design pattern \emph{describes} the resolution of forces in the  target domain; at least  in the setting we're interested in, creating a new design pattern also \emph{effects} a resolution of forces directly.  This The  use case of design pattern development maps into our diagram of the basic features of serendipity as follows:  \input{pattern-schematic-tikz.tex}  diff --git a/serendipity.pdf b/serendipity.pdf  index a0a88db..824d845 100644  Binary files a/serendipity.pdf and b/serendipity.pdf differ