<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Christopher Jones on Medium]]></title>
        <description><![CDATA[Stories by Christopher Jones on Medium]]></description>
        <link>https://medium.com/@christopherjones_93188?source=rss-7060f0523ade------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*4Q810VPNSvOvZzvHidYd-w.jpeg</url>
            <title>Stories by Christopher Jones on Medium</title>
            <link>https://medium.com/@christopherjones_93188?source=rss-7060f0523ade------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 01:48:36 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@christopherjones_93188/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Your Software Is Not Built: It Emerges]]></title>
            <link>https://medium.com/@christopherjones_93188/your-software-is-not-built-it-emerges-4c1f06a4c5b0?source=rss-7060f0523ade------2</link>
            <guid isPermaLink="false">https://medium.com/p/4c1f06a4c5b0</guid>
            <category><![CDATA[software]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[physics]]></category>
            <category><![CDATA[code]]></category>
            <dc:creator><![CDATA[Christopher Jones]]></dc:creator>
            <pubDate>Mon, 27 Nov 2017 03:58:46 GMT</pubDate>
            <atom:updated>2017-11-27T03:58:46.714Z</atom:updated>
            <content:encoded><![CDATA[<p>I was recently reading a wonderful book, <a href="http://i%20hope%20you%20enjoyed%20this%20article.%20if%20you%20like%20it%2C%20please%20share%20and%20follow%20me%21/">The Big Picture by Sean Carrol</a>l, and even though it was about physics and philosophy, it had me thinking about computer science. In the natural sciences, there is a term called emergence. The basic idea is that smaller individual units make up a whole but that whole has entirely different properties than each individual unit. Software, itself, is an emergent phenomenon. Let’s get into it, but first, we must define emergence and then how we describe an emergent phenomenon.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/450/0*nmPkeOeCvoEQA2PK.jpg" /></figure><h3>EMERGENCE</h3><p>To understand emergence consider an insect colony. The insects, themselves, are unique individuals with metabolic and reproductive processes. However, the colony is also a sole entity. An ant colony can swarm a dropped cookie, a bee colony can build enormous hives, and a termite colony can build great cathedrals.</p><p>While colonies are somewhat organized by giving roles to individuals within the colony, seemingly disorganized swarms of insects such as bark beetles or gypsy moths can destroy several acres of forest per day.</p><p>It is not that the whole is greater than the sum of its parts but that the whole is different from individual parts. An individual cell that operates according to the strict laws of biochemistry is nothing like the complex, organized organism that arises. A single change to a single cell will not affect the larger organism but similar single changes in many individual cells will. In this way, the collective properties are not disturbed by individual outliers but by real shifts in information.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/439/0*DG5ZI6op8EM3fh1M.jpg" /></figure><p>An individual cell (such as cancer) that manages to influence other cells around it becomes a danger in preserving the collective whole. So let’s bring this idea of emergence into the everyday human domain.</p><h3>MODELS</h3><p>The stock market, the wave in sports stadiums, and the Billboard Hot 100 music chart are examples of individual units producing a collective whole. The internet itself is a vast network of individual nodes. When Adam Smith, the famed economist, claimed that an”invisible hand” would kick in to regulate free markets according to collective desires of consumers, he was referring to an emergent phenomenon. As we compile data of collective behavior, we instinctively try to make sense of it all, and in doing so we assign definition and form. We use mental models. These models are constructed in language that we humans can understand and in forms that we humans can conceptualize.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/0*8YnsNVd5xMo8HLFB.png" /></figure><p>Significantly, an important aspect of human-constructed models is how we humans describe them. They are not real in a physical sense, yet they exist to us because they are relevant to our experience as human beings. Our models must conform to the context they are being used to describe. To illustrate this, you cannot touch an economy, but you can touch something that exists as a result of that economy. Or, you don’t describe your friend as a collection of atoms, but simply as your friend. Describing everything as just collections of atoms is not wrong, but it is simply not a useful enough way to talk about the world.</p><p>Consequently, we use tools of measurement and observation to craft our models. Because we know that with enough information to build our models, these ideas, these seemingly fake things, can be manipulated and controlled with very real consequences. They are relevant because they are real to us. There are limits to what we can model, notably, in that they remain mostly unfinished. In order for a model to be accepted, they only need to be accurate enough to explain something in a manner we find sufficient and remain in the context of its domain. Having said that, this is where we come to software.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Lj5phYr008F1RF-s.jpg" /></figure><p>First, it is insufficient to say that a software application does not exist before any code is written. In software engineering, a requirements specifications document serves as a description of an intended piece of software. Software engineers use specifications to design the underlying components to produce software. That is to say, the document itself is a model OF a model. The potential for all software to be written is always there because the code and underlying components are always there. The model emerges as the specification requirements are written. Similarly, for programmers that do not use specification requirements documents, but just hammer in code, there is a model in their heads of what they want their code to do. A model exists before any code is in place. It is real. It can inspire action and motivate individuals.</p><p>So, what happens as we build our application?</p><h3>INCOMING SOFTWARE</h3><p>The application emerges into form piece-by-piece as the code is written. The software is checked against the requirements. Again and again, engineers ask themselves does it sufficiently match the model? Recall that the application isn’t real, but the models for it are.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*YP50crKND4R7Y6Wi.png" /></figure><p>As developers, we do our best to build on top of dependable operating systems, to use trusted frameworks, to establish coding standards, and to use reliable network protocols. Using a small amount of code to accomplish a task gives an engineer a sense of pride. Still, even the smallest program must be tested before deployment. Just like our models of nature and human behavior, we cannot always predict the result of software. This is partly why Computer Science is a science. More often than not, you have to experiment; you run the compiler and await the results.</p><h3>STICKING TO THE MODEL</h3><p>The problem is, again, our models are imperfect. Unlike the earlier examples of nature and society, the engineer has to achieve a model that does not yet exist. It was guessed to be possible using calibrations, domains, and restrictions in a process of brainstorming by the creator using what they believe could be built. And, like how we model nature and society, our models of software are not exact.</p><p>Out of the written code emerges unintended consequences. With building a house, stacks of wood become a wall, and in software, stacks of methods become a module. But, there are important differences. A wall has few outcomes; it typically acts as intended or fails spectacularly. A program has many moving parts, in the form of underlying logic gates, and those moving parts can affect the system from the steel on up to the end user. Each of these parts is defined by the code. The written portion of the application must conform to the different layers of moving parts in its host computer and also within itself. The database is a model. The workflow and way the code is deployed is a model. There are many other models underlying the application. Markedly, the most finicky of those models is the code structure, itself.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*tom70rjTgH3C0kPA.jpg" /></figure><h3>THE LEGACY EMERGES</h3><p>In a final twist to add further complexity to the system, the code itself is written with some structure in mind. This structure is yet another model. With enough time and unexpected bugs, the increasing maintenance cost of the code eventually threatens to overcome the business value of the system. As new features are added, the code becomes more complicated. Parts of the code are not documented. New parts of the code do not match code standards. New modules are built in a different style. The model of the code eventually falls apart, and if not refactored, another final emergent property tends to make itself known. This property has many names in the industry (including expletives) but in 1992, computer scientist Ivar Jacobson famously called it software entropy.</p><p>I hope you enjoyed this article. If you like it, please share and follow me!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4c1f06a4c5b0" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>