deletions | additions       diff --git a/bibliography/biblio.bib b/bibliography/biblio.bib  index 4d4938c..790f37d 100644  --- a/bibliography/biblio.bib  +++ b/bibliography/biblio.bib 
...
@article{meszaros1998pattern,  title={A pattern language for pattern writing},  author={Meszaros, Gerard and Doble, Jim},  journal={Pattern languages of program design},  volume={3},  pages={529--574},  year={1998}  }  @inproceedings{schmidhuber2007simple,  title={Simple algorithmic principles of discovery, subjective beauty, selective attention, curiosity \& creativity},  author={Schmidhuber, J{\"u}rgen},  diff --git a/future-work.tex b/future-work.tex  index f33e27b..c0c634f 100644  --- a/future-work.tex  +++ b/future-work.tex 
...
\emph{invention step}, which combines the classified trigger  $T^{\star}$, and preparations $p^{\prime}$, and produces a novel  result $R$ -- to be \emph{blends} in the sense of Joseph Goguen  \citeyear{goguen1999introduction}. The epistemological framework of discovery gives some important clues  about how to compute a common base between $T$ and $p$, a key step for  blending. blending, since these common features will typically be preserved in  the blend.  Although $T$ was previously uninteresting, it will have attributes or attribute-types that match the patterns recognised by  $p$ (e.g. van Andel's \citeyear{van1994anatomy} ``\emph{One \emph{One  surprising observation}''). observation}).  %  In the invention step, reasoning, experimentation, social interaction  strategies rely on $p^{\prime}$, which might draw on patterns like van  Andel's \emph{Successful error} in order to pinpoint the seeds of a useful result  within $T^{\star}$. One important guidepost for implementation is  the theory-building orientation that says that outcomes should sometimes result in may include  new patterns of behaviour that the system can draw on in subsequent interactions.  What is particularly needed is an approach to encoding patterns and 
...
\begin{itemize}  \item[(1)] We want to encode our design patterns directly in runnable  programs, not just give them to programmers as heuristic guidance.  \item[(2)] We want the automated programming system (automated) programmer  to generate new design  patterns, not just apply or adapt old ones.  \item[(3)] We want our design patterns to help describe new problems,  not just capture the solutions to existing problems.  \end{itemize}  \citeA{meszaros1998pattern} describe the typical scenario for design  pattern writers: ``You are an experienced practitioner in your  field. You have noticed that you keep using a certain solution to a  commonly occurring problem. You would like to share your experience  with others.'' They also remark that ``What sets patterns apart is  their ability to explain the rationale for using the solution (the  `why') in addition describing the solution (the `how').'' Regarding  the criteria that pattern writers seek to address, they write: ``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 at the meta-level of pattern writing, e.g. \emph{Clear target  audience}, \emph{Visible forces}, and \emph{Relationship to other  patterns}. A good pattern describes the resolution of forces, but  it also resolves certain forces itself. In terms of our now-familiar  diagram:  \input{pattern-schematic-tikz.tex}  This diagram does not suggest that every instance of ``a solution to a  problem in a context'' is serendipity -- on the contrary, that is just  the discovery step. To van Andel's assertion that ``The very moment I  can plan or programme `serendipity' it cannot be called serendipity  anymore'' -- of course, if the context is fully determined in advance,  and if the solution is completely replicable, then some of the  fundamental conditions for serendipity are not met. However, we can  also describe patterns with built-in indeterminacy:  \begin{mdframed}  \vspace{-.35cm}  \paragraph{Successful error}  \emph{Van Andel's example} -- Post-it\texttrademark\ Notes\\[.05cm]  \begin{description}  \item[{\tt context}] -- You run a creative organisation with several different divisions and many contributors with different expertise. Unexpected discoveries are often made.  \item[{\tt problem}] -- One of the members of your organisation  discovers something with interesting properties, but that no one  knows how to turn into a product with industrial or commercial application.  \item[{\tt solution}] -- You create a space for sharing and discussing  interesting ideas on an ongoing basis (perhaps a Writers Workshop).  \item[{\tt rationale}] -- You suspect it's possible that one of the  other members of the firm will come up with an idea about an  application; you know that if a potential application is found, it  may not be directly marketable, but at least there will be a  prototype that can be concretely discussed.  \item[{\tt resolution}] -- Writing down and promulgating the  \emph{Successful error} pattern using this template gives one such  prototype.  \end{description}  \end{mdframed}  diff --git a/serendipity.tex b/serendipity.tex  index 0170569..f4aad47 100644  --- a/serendipity.tex  +++ b/serendipity.tex 
...
\usepackage[T1]{fontenc}  \usepackage[utf8]{inputenc}  \usepackage{setspace}  \usepackage[framemethod=tikz]{mdframed}  \mdfsetup{  skipabove=\baselineskip,