<?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 Martin Breuss on Medium]]></title>
        <description><![CDATA[Stories by Martin Breuss on Medium]]></description>
        <link>https://medium.com/@martin.breuss?source=rss-9ec36671f354------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*q3j6lqMGB5Livx2exJIuVg.png</url>
            <title>Stories by Martin Breuss on Medium</title>
            <link>https://medium.com/@martin.breuss?source=rss-9ec36671f354------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 12:49:18 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@martin.breuss/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[All you need for working with Python 3 on a Mac]]></title>
            <link>https://medium.com/@martin.breuss/all-you-need-for-working-with-python-3-on-a-mac-c811ac094869?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/c811ac094869</guid>
            <category><![CDATA[beginner-coding]]></category>
            <category><![CDATA[setup]]></category>
            <category><![CDATA[beginners-guide]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[python3]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Sat, 08 Sep 2018 18:50:16 GMT</pubDate>
            <atom:updated>2018-09-08T18:51:01.122Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FyEkUvuwlT7IGHVtmD9sjA.png" /></figure><p>To clear up any confusion about a functioning setup to start working with Python 3 on a fresh Mac — here’s a quick (minimal) guide.</p><h3>DO #1: download and install Python 3.x.x</h3><p>Go to <a href="https://www.python.org/downloads/">the official website</a> and download Python from there. Yes — since you’re asking — <strong>do get the latest version</strong> (the one with the shiny yellow button):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*X5oLo0KLz4XZ6YQc9y5psA.png" /></figure><p>Execute the installer and step through the required steps…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/549/1*3Q0aCaGwIIZvZjo5lwCSDg.png" /><figcaption>(it’s easier than that)</figcaption></figure><p>And <strong>if you’re on Mavericks (10.9) or above, you’re (nearly) done</strong>.</p><h4>&gt;&gt; Really? Proof it!</h4><p>Open your Terminal and verify by typing:</p><pre>python3 --version</pre><p>which should respond to you with something similar to: <em>Python 3.7.0</em></p><p>Then type:</p><pre>git --version</pre><p>which should either:</p><ul><li>prompt you to install the command-line tools. If so, allow that install (it’ll bring version control to your life!). <em>This is, btw, the step that completes the “(nearly)” from above.</em></li><li>or, if it’s already there, it will say something like: <br><em>git version 2.15.2 (Apple Git-101.1)</em></li></ul><p>Finally, you can also ask your patient terminal the following:</p><pre>pip3 --version</pre><p>and it will dutifully respond with a lengthy: <em>pip 18.0 from /Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/pip (python 3.7)</em></p><p>And with that you are all set up and ready to dive into <strong>writing code with Python 3</strong>, and <strong>installing whatever package</strong> you may fancy from Python’s rich ecosystem on <a href="http://pypi.org/">pypi.org</a> using <strong>pip3</strong> (<a href="https://medium.com/@martin.breuss/virtual-environments-for-python-cc8f18a48c6c">create a <strong>virtual environment</strong></a><strong> first</strong>!), and indulge in the reassuring beauty of <strong>version control</strong> with <strong>git</strong>.</p><h4>&gt;&gt;&gt; What if your OS is older than 10.9?</h4><p>Update it. It’s free and makes your computer better.</p><p>If for any reason you can’t update your OS but still want to proceed working with Python 3 and git, then <a href="https://git-scm.com/book/en/v2/Getting-Started-Installing-Git">get Git from here</a>:</p><blockquote>A macOS Git installer is maintained and available for download at the Git website, at <a href="http://git-scm.com/download/mac">http://git-scm.com/download/mac</a>.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*DseohaM183bB7_Zh" /></figure><blockquote>You made it! Good job! Now, for your enjoyment, here’s a list of steps that are often mentioned in tutorials on how to get started, but IMO they only complicate the basic setup.</blockquote><h3>DON’T #1: install Anaconda</h3><p>Anaconda is a great <strong>tool for data analysis</strong>, however, it comes with <a href="https://stackoverflow.com/a/42096429/5717580">a lot of stuff</a>. Anaconda is a full-blown Python <em>distribution</em> (= it’s its own <em>version</em> of Python!), that comes with its own package manager, tons of data science packages pre-installed, and oh — did I mention it also includes <em>R</em>?!</p><h3>DON’T #2: install homebrew</h3><p>Every second tutorial tells you to install homebrew first, and then use it to get python on your mac. Don’t do it. There’s no need.</p><p>I personally don’t like cellars very much, and even though I tried to homebrew a couple of years ago (I only drank it for the experience, not the taste ;) that alone can’t win me over for the easily forgettable and potentially already long-forgotten leftovers that homebrew’s <em>Cellar</em> structure induces on my machine.</p><h3>DON’T #3: install pip</h3><p>You <strong>don’t have to force-install pip</strong> if you download Python 3 from the official download page. It comes along voluntarily.</p><p>the <a href="https://pip.pypa.io/en/latest/installing/">official pip site</a> speaks:</p><blockquote>pip is already installed if you are using Python 2 &gt;=2.7.9 or Python 3 &gt;=3.4 downloaded from <a href="https://www.python.org/">python.org</a></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IauQ9ox4OojpggX1sHwGjA.png" /></figure><h4>+/- useful tutorials</h4><ul><li><a href="https://docs.python.org/3/using/mac.html">https://docs.python.org/3/using/mac.html</a></li><li><a href="https://itsevans.com/install-pip-osx/">https://itsevans.com/install-pip-osx/</a></li><li><a href="https://www.digitalocean.com/community/tutorials/how-to-install-python-3-and-set-up-a-local-programming-environment-on-macos">https://www.digitalocean.com/community/tutorials/how-to-install-python-3-and-set-up-a-local-programming-environment-on-macos</a></li><li><a href="https://docs.python-guide.org/starting/install3/osx/">https://docs.python-guide.org/starting/install3/osx/</a></li><li><a href="https://www.stuartellis.name/articles/mac-setup/">https://www.stuartellis.name/articles/mac-setup/</a></li><li><a href="https://realpython.com/python-first-steps/">https://realpython.com/python-first-steps/</a></li></ul><h4>Other Resources</h4><ul><li><a href="https://pip.pypa.io/en/latest/installing/">https://pip.pypa.io/en/latest/installing/</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c811ac094869" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Django WebDev: Keeping secrets with Environment Variables]]></title>
            <link>https://medium.com/@martin.breuss/django-webdev-keeping-secrets-with-environment-variables-78def8813cf2?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/78def8813cf2</guid>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Mon, 04 Jun 2018 11:59:48 GMT</pubDate>
            <atom:updated>2018-06-04T11:59:48.838Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*60nAsf4GBYEWiF73MrvPlg.jpeg" /></figure><p>When working in Web Development with e.g. APIs of other services, you will often need to include <strong>Authentication</strong> to access those services. These can come in the form of e.g. passwords, API keys or secrets. We ❤ Version Control, however ( — limitless optimism aside — ), we don’t want to share those personal secrets with the open source community.</p><p>There are horror stories about significant charges being accrued from carelessly posting API secrets to public version control (links at the bottom).</p><blockquote>Bots are quick, and one compromised commit is one too many.</blockquote><p>There are a couple of ways to avoid such inconvenience (see links at the bottom). This article explains how to achieve safety by using <strong>Environment Variables</strong> in UNIX systems. Specifically: <strong><em>Virtual Environment Variables</em></strong>, since I’d suggest everyone to do their webdev work inside of <a href="https://medium.com/@martin.breuss/virtual-environments-for-python-cc8f18a48c6c">Virtual Environments</a>.</p><h3>Part 1: Environment Variables (EVs)</h3><h4>Ok, so what is an Environment Variable?</h4><p><a href="https://en.wikipedia.org/wiki/Environment_variable">Environment variables</a> are variables that are accessible from anywhere inside your local system and are usually shared by multiple applications. In UNIX systems, the most famous one is <a href="https://en.wikipedia.org/wiki/PATH_(variable)">$PATH</a> which specifies file paths where executables are located.</p><h4>How do they live up to this challenge?</h4><p>Once initiated, they are accessible from anywhere within our project <strong>without ever needing to write the actual value down</strong>. We simply refer to that value through calling the environment variable. This allows the convenience of working with API secrets and passwords anywhere in the project, while keeping all these secrets safe to ourselves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/275/1*IYC5w0CUD_2YGt8rMLjDtA.jpeg" /></figure><h3>*Let’s give it a spin* 🎡</h3><p>Open up your CLI and type printenv. This will give you a list of all the current environment variables present on your system. E.g.:</p><pre>SHELL=/bin/zsh<br>HOME=/Users/Martin<br>LOGNAME=Martin<br>USER=Martin<br>PATH=/Library/Frameworks/Python.framework/Versions/3.6/bin:/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/bin:/usr/local/git/bin</pre><p>This is just an excerpt, probably more lines will show up on your machine.</p><p>You can check out the variables’ individual values using echo $&lt;VARNAME&gt; e.g.: echo $LOGNAME returns Martin for me.</p><h4>Adding and removing env-vars</h4><p>You can add a new environment variable with the following command:</p><pre>export &lt;NAME&gt;=&lt;VALUE&gt;</pre><p>Where &lt;NAME&gt; stands for your new env-var&#39;s name and (you guessed it) &lt;VALUE&gt; should be substituted for the value you want to assign.</p><p>Removing EVs is equally easy:</p><pre>unset &lt;NAME&gt;</pre><p>That’s it. Try it out by adding and removing some environment variables using these commands. Remember you can check up on what’s happening using printenv or echo &lt;VARNAME&gt;.</p><h3>Part 2: Virtual Environment Variables (VEVs)</h3><h4>Automate and Compartmentalize</h4><p>Since we want to avoid setting anything for our whole system if we don’t need to — especially since in web development we’re usually talking about <strong>project-specific secrets</strong> — it would be better to use such environment variables <em>inside of virtual environments:</em> <strong>Virtual Environment Variables</strong> basically. I’ll call them VEVs for short. Setting this up is possible, and much easier than having to export and then unset each variable every time. Here we go:</p><p><a href="https://medium.com/@martin.breuss/virtual-environments-for-python-cc8f18a48c6c">After creating a Virtual Environment</a>, open the activate script in your favourite text editor. You can find this file in a well-known location: env/bin/activate</p><p>We will edit this script that gets run every time our venv gets activated — which makes it the best place to let the computer know which environment variables we would like to have, and which ones to unset once we exit the virtual environment.</p><p>First we make sure that the new VEVs won’t be sticking around, so we add unset commands to the deactivate command section. This code gets run every time we deactivate our venv:</p><pre>deactivate () {<br>    unset MY_SUPER_SECRET_SECRET<br>    # lots of other code<br>}</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*W466j8uSvapKAeqO-8JKcQ.jpeg" /><figcaption>keep your environments clean</figcaption></figure><p>And in step 2/2 we add an export including our VEVs value to the end of the script, like so:</p><pre>export MY_SUPER_SECRET_SECRET=&#39;OMG so secret I can&#39;t tell you here!&#39;</pre><p>Safe the script and close it. Now spin up your venv using</p><pre>source env/bin/activate </pre><p>from your home folder, and try toprintenv:</p><p>Behold! MY_SUPER_SECRET_SECRET shows up <strong>including it&#39;s assigned value</strong>.</p><p>Deactivate the venv and try the same thing again. Your secret should be gone. Safe within its comfy virtual environment.</p><h4>Using VEVs</h4><p>To use a VEV inside of your python project, you’ll need to call it using the os module from the standard library:</p><pre><strong>import</strong> os<br>secret = os.environ[&#39;MY_SUPER_SECRET_SECRET&#39;]<br>print(secret)</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6hE7v9yVeto8LgF6C0R6GQ.png" /></figure><h3>Conclusion</h3><p>Some secrets are meant to stay secret, or at least locally on your machine. Using environment variables — and more specifically virtual environment variables — allows us to easily access values within a project without running into the danger of accidentally committing them to public version control. Of course, you’ll have to .gitignore that whole env folder, or else… 😱</p><p>If you’re interested in learning python web development from the ground up and drilling best practices right from the get-go — and if you’ve always wanted to see <a href="https://www.zoobarcelona.cat/en/animals/indian-rock-python">a live Python in the Barcelona zoo</a> (because who wouldn’t want that?!) — take a look at the <a href="http://codingnomads.co/">CodingNomads</a> <a href="http://codingnomads.co/courses/learn-python-barcelona/">python course I’ll be co-teaching this summer in Barcelona</a>. No snake-oil. Just intense quality coding time. :)</p><figure><a href="http://codingnomads.co/"><img alt="" src="https://cdn-images-1.medium.com/max/391/1*Mwwiz_1t8ddrrnWzkZs0OA.png" /></a></figure><p><strong>Horror Stories 😱</strong></p><ul><li><a href="https://securosis.com/blog/my-500-cloud-security-screwup">https://securosis.com/blog/my-500-cloud-security-screwup</a></li><li><a href="https://www.theregister.co.uk/2015/01/06/dev_blunder_shows_github_crawling_with_keyslurping_bots/">https://www.theregister.co.uk/2015/01/06/dev_blunder_shows_github_crawling_with_keyslurping_bots/</a></li><li><a href="https://www.quora.com/My-AWS-account-was-hacked-and-I-have-a-50-000-bill-how-can-I-reduce-the-amount-I-need-to-pay">https://www.quora.com/My-AWS-account-was-hacked-and-I-have-a-50-000-bill-how-can-I-reduce-the-amount-I-need-to-pay</a></li></ul><p><strong>Further Reading (includes alternative paths to safety):</strong></p><ul><li><a href="https://medium.freecodecamp.org/how-to-securely-store-api-keys-4ff3ea19ebda">https://medium.freecodecamp.org/how-to-securely-store-api-keys-4ff3ea19ebda</a></li><li><a href="https://mikegerwitz.com/papers/git-horror-story">https://mikegerwitz.com/papers/git-horror-story</a></li><li><a href="https://softwareengineering.stackexchange.com/questions/205606/strategy-for-keeping-secret-info-such-as-api-keys-out-of-source-control">https://softwareengineering.stackexchange.com/questions/205606/strategy-for-keeping-secret-info-such-as-api-keys-out-of-source-control</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78def8813cf2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Virtual Environments for Python]]></title>
            <link>https://medium.com/@martin.breuss/virtual-environments-for-python-cc8f18a48c6c?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/cc8f18a48c6c</guid>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Thu, 31 May 2018 11:33:09 GMT</pubDate>
            <atom:updated>2018-05-31T11:33:09.641Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VgjNhohCLk56CtIq3Ru0Rg.png" /></figure><p>With a bustling activity as a developer, when grinding projects feels like a morning jog, it is easy for <em>something </em>to go out of hand: <strong>dependencies</strong>.</p><h3>What are dependencies?</h3><p>When building python projects and with the rich ecosystem of modules and packages out there, you’ll want and <em>use all that code</em>. <strong>Dependencies</strong> are python packages containing code that someone else wrote and that you are utilizing for your project. The developers who wrote the code put a lot of effort into making the wheels they built easy to use. Exactly: so that you won’t have to re-invent them.</p><h3>The challenge</h3><p>Re-using code is a great thing! However, fast-forward into maintaining three different Django projects for different clients who are all using <em>different versions of Django </em>and life starts looking less like a paved road. Your system can only have one version of Django installed, so you’ll quickly run into troubles. The examples of such issues with dependencies are manyfold.</p><h3>The solution: virtual environments</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WsFIAvRWP789_nqWO-4LJQ.png" /></figure><p>Maybe you’ve heard about <em>virtual machines</em> before. A virtual machine allows you to run e.g. Linux on your Mac operating system. Programs such as virtualbox help you to create a <strong>separate environment</strong> that is self-sufficient and doesn’t need to interact with the main OS.</p><p><strong>Virtual environments</strong> for python behave similarly, only that they don’t box up a whole OS, but instead:</p><ul><li>a full python installation (including pip)</li><li>whichever modules and packages you add</li></ul><p>Using virtual environments makes software development safer and allows you to keep your handsome hairline intact years longer into the future. Many beginner courses don’t introduce the concept of virtual environments until much later, but learning about them and using them from the get-go can be a huge advantage and time-saver. And it’s pretty easy! So here’s a quick guide for creating our first virtual environment.</p><h3>Creating a virtual environment</h3><p>python 3 makes creating virtual environments (venvs) a breeze. Here’s a quick look at the syntax:</p><pre>python3 -m venv your_venv_name</pre><p>Easy, right?</p><p>By default we’ll call our venv’s simply env - but if you want to go wild with naming them, go ahead. Just remember what you called it!</p><p>Navigate to the folder that was created by this command. Check out which files you can find in /bin. Every freshly minted venv comes with a shiny new version of python as well as pip preinstalled.</p><h3>Using a virtual environment</h3><p>Fine, so all of this env stuff now exists — but behold! You are not yet <em>inside</em> the safe space! In order to enter your new venv you’ll have to source it by typing the magic command:</p><pre>source env/bin/activate</pre><p>All that this does is running the activate script which grants you access to the safe kingdom of venv.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*QmRaoBj1DTtrsin8BccejQ.png" /><figcaption>Mufasa: “already today all these pixels are yours”</figcaption></figure><p>After the command has executed you’ll see an indicator that you are now inside of the virtual environment through having the name you gave the venv in brackets in front of each terminal command.</p><p>For me this looks like so:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/195/1*tset-5cq2xDlz7KbkTMRTw.png" /></figure><p>Depending on the shell you are using it might look a little different (e.g. less colourful) — but the concept is the same.</p><p>Seeing your virtual environment’s name in brackets means you can now start breathing again. You are in a safe place. ;)</p><h3>And… how can I get out of here?</h3><p>To return to your normal system’s environment you only need to say one word to the console:</p><pre>deactivate</pre><p>The bracket-part will disappear and your shell will show your normal prompt again. That’s it. Welcome home. 🏡</p><h3>Quick task: Create a venv and install Django</h3><ol><li>using the terminal create a new folder called testenv</li><li>navigate into the folder</li><li>create a venv called env inside of the foldertestenv</li><li>activate env</li><li>use pip to install Django</li><li>open the python interpreter and import django to verify that it all worked</li></ol><p>🎉</p><blockquote><strong>Note:</strong> Make sure that your virtual environment’s path doesn’t include any whitespace. Otherwise the executables won’t be found and you might end up accidentally using the system-wide install of pip. See: <a href="https://stackoverflow.com/a/7911058/5717580">https://stackoverflow.com/a/7911058/5717580</a> for more information.</blockquote><p>If you want to learn more about Web Development with python and Django (and everything that comes with it) feel free to contact me about the <a href="http://codingnomads.co/courses/learn-python-barcelona/">python bootcamp I’ll be co-teaching this summer in Barcelona</a> with <a href="http://codingnomads.co/">CodingNomads</a>! Maybe you’ll look for your private safe little kingdom in Europe on the way ; )</p><figure><a href="http://codingnomads.co/"><img alt="" src="https://cdn-images-1.medium.com/max/391/1*Mwwiz_1t8ddrrnWzkZs0OA.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cc8f18a48c6c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Remembering the forgetting curve]]></title>
            <link>https://medium.com/@martin.breuss/remembering-the-forgetting-curve-d88ad082502e?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/d88ad082502e</guid>
            <category><![CDATA[learning]]></category>
            <category><![CDATA[remembering]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[spaced-repetition]]></category>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Mon, 21 May 2018 11:27:57 GMT</pubDate>
            <atom:updated>2020-01-17T12:09:19.095Z</atom:updated>
            <content:encoded><![CDATA[<p>🔔 — <em>Ebbinghaus wants to send you push notifications — </em>🔔</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jGLf3peHNmyCbgb3IFbPpw.png" /></figure><p>Everyone knows this existential angst when realizing that {xyz}, which we learned only {xyz} time units ago, is <em>already</em> blanked off our mind.</p><p>How do some people just remember everything?</p><p>Turns out that it’s just about <strong>consistent work</strong> (as with most things in life). Being awesome at remembering is <em>actually not that difficult</em>.</p><p>So… where do all those ________ spots in our recently polished minds come from, and what can we do to fill them up?</p><h3>Probing the leaking mind</h3><p>In <a href="https://en.wikipedia.org/wiki/Hermann_Ebbinghaus#Contributions_to_memory">____</a>, a <a href="https://en.wikipedia.org/wiki/Kingdom_of_Prussia">_____</a> <a href="https://en.wikipedia.org/wiki/Psychology">______</a> named <a href="https://en.wikipedia.org/wiki/Hermann_Ebbinghaus">________</a> Ebbinghaus did his research and came up with data for a very interesting and influential graph called:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZNz28F-L7n0N2A5ums5oqg.png" /></figure><p>When seeing this graph for the first time, a lot of people are mildly shocked.</p><blockquote>That’s <strong>a lot</strong> of forgetting and it happens <strong>very quickly</strong>.</blockquote><p>But everyone has taken a ride on the steep and speedy slope of the forgetting curve, so there’s nothing to deny. Thanks for drawing it out, Ebbi!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/415/1*ggw9fkdQA3jedw_d3YV8hA.png" /></figure><p>But don’t despair — there’s green lines on the horizon (see the graph above)! As humans, we rarely learn something once and then <em>just</em> <em>know it</em>. That’s not how our minds are made up — and for good reasons! Imagine <em>forever</em> remembering everything that you’ve thought or encountered <em>only one time</em>.<br>😮 =💥. Literally.</p><p>So that’s why we sift and discard.</p><p>It’s only worth to remember something when it keeps affecting you <em>more than just once</em>. When something comes around <em>again and again</em>, then it’s probably worth to get out the permanent brain marker and scribble a note onto our spongy internal whiteboard.</p><h3>Spaced Repetition</h3><p>What Ebbinghaus famously investigated was the <strong>spacing</strong> of <strong>repetition</strong> that yields optimal results for <strong>remembering</strong>. It essentially goes like this:</p><blockquote>At first you need to revisit more frequently, then in ever more receding time intervals. At some point the information is there to stay.</blockquote><p>So: studying flashcards actually works — if you keep the <strong>intervals</strong> in mind. Some smart people developed <a href="https://en.wikipedia.org/wiki/Spaced_repetition#Algorithms">a few algorithms</a> to make the tedious task of remembering to remember to study in order to remember — <em>haaaaaa! deep breath!</em> — much easier.<br>If you want to give it a try, <a href="https://apps.ankiweb.net/">ANKI</a> is very powerful and relatively easy to use. Keep in mind, though, that this is <em>just for remembering</em> — in order to learn something, you’ll need to <a href="https://www.supermemo.com/en/articles/20rules">understand it first</a>.</p><p>I did a few experiments using spaced repetition systems for studying code. I learned a lot in the process of creating the flashcards, and stuck some essentials to my biological neural nets through revisiting with a spaced-repetition system. It’s very powerful for learning new languages as well, and a staple for med students who need to memorise large amounts of information.</p><h3>Reflection for learning 🎉</h3><p>What I personally find interesting, is how it ties into the suite of <strong>reflective actions</strong> that make learning much more effective:</p><ul><li>feedback</li><li>self-review (reflection)</li><li>repetition (practice)</li><li>mentoring</li></ul><p>I’ve experienced all the above points as crucial factors for learning and improving — both on a professional as well as on a personal level.</p><p>However, I think it’s easy to become lazy about any and all of these four points and stop the learning process after <em>understanding</em> an idea. Sometimes even before that, especially when it’s about studying for a test… 😳</p><p>So I find it important to remind myself that <strong>actually learning something is more than understanding it</strong>. It doesn’t stop once you <em>got it</em>. There’s work that needs to be done to get to that high point of 🤔<em>AHA!</em>🤓, and there’s more work to be done to move the 🤔<em>AHA!</em>🤓 into permanent actionable memory.</p><p>Spaced repetition systems, such as ANKI, are a great tool to help remembering pieces of information, and can successfully be used to counteract the forgetting curve.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/402/1*GjAyt_DxqX4-9CpvbfvCkg.png" /><figcaption>So maybe next time I’ll remember : )</figcaption></figure><p>If you found this article worthwile, <strong>please give me a clap</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/87/0*dk_oy7-tb72-kuE4." /></figure><p>This helps me understand that I’m doing something useful for others. :)</p><p>And if you are interested in learning python programming with me while traveling the world, check out <a href="https://codingnomads.co/">my company</a> where we do just that.</p><figure><a href="http://codingnomads.co/"><img alt="" src="https://cdn-images-1.medium.com/max/391/1*Mwwiz_1t8ddrrnWzkZs0OA.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d88ad082502e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Customising Sublime Text for Python Web Development]]></title>
            <link>https://medium.com/@martin.breuss/customizing-sublime-text-for-python-web-development-de31873d2d06?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/de31873d2d06</guid>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[optimization]]></category>
            <category><![CDATA[sublimetext]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[web-development]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Sat, 12 May 2018 13:45:31 GMT</pubDate>
            <atom:updated>2018-10-01T19:49:56.360Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JX_aeQBdZnbHB19EYM-Vvw.png" /></figure><p>Sublime Text’s <strong>rich package ecosystem</strong> makes it possible to fine-tune the text editor specifically to what you are using it for. Some of the packages are universally helpful, others, such as syntax checking, are specific to the programming language you are working with.</p><p>Changing settings is easy and might also help to clarify how to do further customisations — if that is where your ❤️ is at.</p><h3>Sublime Settings</h3><p>Before moving onto installing plugins, there are a couple of useful settings that can help to tailor Sublime for python development. Go to <strong>Sublime Text → Preferences → Settings</strong> and in the <em>Preferences.sublime.settings-User</em> file enter the following rules:</p><pre>{<br>    &quot;spell_check&quot;: false,<br>    &quot;rulers&quot;:<br>    [<br>        72, 79<br>    ],<br>    &quot;tab_size&quot;: 4,<br>    &quot;translate_tabs_to_spaces&quot;: true,<br>    &quot;trim_trailing_white_space_on_save&quot;: true,<br>    &quot;ensure_newline_at_eof_on_save&quot;: true<br>}</pre><p>I personally prefer to keep &quot;spell_check&quot; on false because I don’t like to see red lines everywhere when writing non-dictionary-compliant words. Maybe I’m just damaged by the school system. If you have a different opinion on this, simply set &quot;spell_check&quot; to true.</p><p>The magic numbers for the &quot;rulers&quot; setting originate from pythons PEP8. Docstrings and comments should be limited to 72 characters and all other lines should not be longer than 79 characters. The ruler lines can help to remind you when you are approaching these limits.</p><p>In good python fashion we set the &quot;tab_size&quot; to 4 and &quot;translat_tabs_to_spaces&quot; to <a href="https://www.youtube.com/watch?v=SsoOG6ZeyUI">avoid relationship troubles</a>.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FSsoOG6ZeyUI%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DSsoOG6ZeyUI&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FSsoOG6ZeyUI%2Fhqdefault.jpg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/6b753341dc57e39f534315c648060739/href">https://medium.com/media/6b753341dc57e39f534315c648060739/href</a></iframe><p>Finally, we can get another PEP8 compliance automation with &quot;ensure_newline_at_eof_on_save&quot; which does just that: adding a single newline at the end of the file when you save the document.</p><h3>Packages (Plugins)</h3><p>And now we dive right into the pool of great packages for Sublime Text that extend its functionality and can make your life as a programmer easier and more colorful!</p><h4><a href="https://github.com/SideBarEnhancements-org/SideBarEnhancements">SidebarEnhancement</a></h4><p>Power-up for the ST sidebar, introducing useful features such as being able to create new files directly from there and sending files to the recycle bin when they are deleted from inside Sublime.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/244/1*Du8DeCowEFtdwQLAjV0OKQ.jpeg" /></figure><h4><a href="http://damnwidget.github.io/anaconda/">Anaconda</a></h4><blockquote><strong>Note: </strong>this is <em>not </em>the popular python distribution often used for data science that goes by the same name.</blockquote><p>The Sublime Text Anaconda plugin helps to make ST more IDE-like while keeping it still operating much quicker than most IDEs. It introduces easy access to useful features such as:</p><ul><li>Code autocompletion</li><li>Code linting: syntax errors</li><li>Code linting: highlight and automatically fix PEP8 violations</li><li>Python documentation on class, method or function under the cursor</li><li>Goto Python definition anywhere in the project</li><li>Find object usage across the project</li><li>and lots more…</li></ul><h4><a href="https://packagecontrol.io/packages/AutoDocstring">Auto Docstring</a></h4><p>Auto Docstring does just what its name suggests — it allows you to swiftly create a Python docstring template for your selected function, method or class. You can even add docstrings for every function in a file and change between Google and Numpy style even after the docstrings were created. A very basic auto-generated example would look like this:</p><pre>def my_function():<br>    &quot;&quot;&quot;Summary<br>    <br>    Returns:<br>        TYPE: Description<br>    &quot;&quot;&quot;<br>	return None</pre><h4><a href="http://www.sublimelinter.com/en/stable/">SublimeLinter</a></h4><p>For python syntax and PEP8 you could stick with using the linting provided through the Anaconda plugin. However, using <strong>SublimeLinter</strong> has the advantage that it is extensible also with other linters that are useful for web development. That’s why I would suggest to disable Anaconda’s linter, like so:</p><p>Use the file menu: <strong>Sublime → Preferences → Package Settings → Anaconda → Settings — User </strong>to go to the user-defined Anaconda settings file. Then set: {&quot;anaconda_linting&quot;: false}</p><p>Now, install <strong>SublimeLinter</strong> using the Package Manager.</p><blockquote><strong>Note: </strong>Most of the linters require additional packages to operate correctly, so make sure to read the instructions carefully</blockquote><p><a href="https://github.com/SublimeLinter/SublimeLinter-pyflakes"><strong>SublimeLinter-pyflakes</strong></a><strong><br>pyflakes</strong> performs syntax checking on python files. This package needs to be installed using pip before adding it to ST with the Package Manager</p><p><a href="https://github.com/PyCQA/pycodestyle"><strong>SublimeLinter-pycodestyle</strong></a><strong><br>pycodestyle</strong> is the new name of the pep8 package. Install it using pip, then add it to Sublime using the Package Manager for PEP8-compliant code style linting.</p><p><strong>Other useful linters for web development:</strong></p><ul><li><a href="https://sublime.wbond.net/packages/SublimeLinter-json">SublimeLinter-json</a></li><li><a href="https://sublime.wbond.net/packages/SublimeLinter-jshint">SublimeLinter-jshint</a></li><li><a href="https://sublime.wbond.net/packages/SublimeLinter-csslint">SublimeLinter-csslint</a></li><li><a href="https://sublime.wbond.net/packages/SublimeLinter-html-tidy">SublimeLinter-html-tidy</a></li><li><a href="https://sublime.wbond.net/packages/SublimeLinter-pyyaml">SublimeLinter-pyyaml</a></li></ul><h4><a href="https://github.com/squ1b3r/Djaneiro">Djaneiro</a></h4><p>This useful plugin adds a couple of handy Django snippets with tab autocompletion into ST to make working on web projects quicker.</p><p>E.g. block + TAB generates {% block %}{% endblock %}</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/278/1*_D7FGqRxT_tIUw7wLEeqcg.gif" /></figure><p>Very cool. Check out all it has to offer <a href="https://github.com/squ1b3r/Djaneiro">here</a>!</p><h4><a href="https://github.com/kemayo/sublime-text-git/wiki">Git</a></h4><p>Allows to do the most common set of git commands from within Sublime. <br>git add, commit, push etc. right from your favorite text editor.</p><h4><a href="https://github.com/jisaacks/GitGutter">GitGutter</a></h4><p>GitGutter helps to keep track of which files have been edited since the last commit. It adds small symbols next to the line numbers that indicate their status in relation to the git repository.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/349/1*LfrYvdN2-0vx_uDLWGN9sw.png" /></figure><h4><a href="https://github.com/wuub/requirementstxt">requirementstxt</a></h4><p>Introduces syntax highlighting for <em>requirements.txt</em> files. Be good to your 👀!</p><h4><a href="https://facelessuser.github.io/MarkdownPreview/">Markdown Preview</a></h4><p>Allows to preview Markdown documents in your browser — build them using different available flavors!</p><h4>Other useful Web Dev Packages</h4><p><a href="https://github.com/y0ssar1an/CSS3"><strong>CSS3</strong></a><strong><br></strong>The CSS3 plugin allows for futuristic CSS syntax highlighting. You’ll need to tweak the rest of the settings a bit, but it’s easy to follow along the steps here:</p><p><a href="https://github.com/Monnoroch/ColorHighlighter"><strong>ColorHighlighter</strong></a><strong><br></strong>This one is fun. Adds more 🌈 to your code. Hex color values will be displayed with the corresponding color as background. Makes it easier to remember what color you were adding where.</p><p><a href="https://emmet.io/"><strong>Emmet</strong></a><strong><br></strong>If you’re excited about automating the repetitive parts of your web development work, then Emmet can be a big help. Install the <a href="https://github.com/sergeche/emmet-sublime">package</a>, then learn how to benefit: <a href="https://docs.emmet.io/cheat-sheet/">https://docs.emmet.io/cheat-sheet/</a></p><p>With adding these packages and tweaking the settings as mentioned above, Sublime Text becomes a lightweight but powerful editor for web development with python.</p><p>If you are just getting started and are looking for a <strong>fun and affordable bootcamp</strong> that helps you to learn programming and gets you to travel and see the world, check out <a href="http://codingnomads.co/">CodingNomads</a> — I teach there :) 💻 🎒🏝</p><figure><a href="http://codingnomads.co/courses/learn-python-barcelona/"><img alt="" src="https://cdn-images-1.medium.com/max/391/1*Mwwiz_1t8ddrrnWzkZs0OA.png" /></a></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/360/1*Mm-yYSOUG_o66JzaGCIeuQ.png" /></figure><h3>More Info:</h3><ul><li><a href="https://medium.com/@arshamshirvani/sync-your-sublime-text-3-configurations-safely-and-easy-b493021c80da">Sync your settings to GitHub</a></li></ul><p><strong>Plugin Setup:</strong></p><ul><li><a href="http://www.gtlambert.com/blog/python-sublime-text-3">http://www.gtlambert.com/blog/python-sublime-text-3</a></li><li><a href="https://www.fullstackpython.com/sublime-text.html">https://www.fullstackpython.com/sublime-text.html</a></li><li><a href="https://realpython.com/setting-up-sublime-text-3-for-full-stack-python-development/">https://realpython.com/setting-up-sublime-text-3-for-full-stack-python-development/</a></li><li><a href="https://nazrul.me/2017/05/17/make-sublime-text-as-the-best-ide-for-full-stack-python-development/">https://nazrul.me/2017/05/17/make-sublime-text-as-the-best-ide-for-full-stack-python-development/</a></li></ul><p><strong>Learn to increase your performance when working with ST3:</strong></p><ul><li><a href="https://hackernoon.com/super-charge-your-sublime-text-3-to-increase-your-productivity-5d02c2c1b356">https://hackernoon.com/super-charge-your-sublime-text-3-to-increase-your-productivity-5d02c2c1b356</a></li><li><a href="https://fosstack.com/setup-sublime-python/">https://fosstack.com/setup-sublime-python/</a></li></ul><p>If you found this article interesting, <strong>please give me a clap</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/87/1*SKDqlIFgIVTygC3sWqENQA.png" /></figure><p>This helps me understand whether I’m doing something useful for others. :)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=de31873d2d06" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Finding a useful Dictionary API]]></title>
            <link>https://medium.com/@martin.breuss/finding-a-useful-dictionary-api-52084a01503d?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/52084a01503d</guid>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[language]]></category>
            <category><![CDATA[search]]></category>
            <category><![CDATA[dictionary]]></category>
            <category><![CDATA[api]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Thu, 09 Feb 2017 03:00:14 GMT</pubDate>
            <atom:updated>2020-02-24T01:59:57.994Z</atom:updated>
            <content:encoded><![CDATA[<p>I thought that would be easy, but they are difficult to spot</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bWkq8BoeR3M07gsn9xPwew.jpeg" /><figcaption><em>“Check out my new dictionary App!”</em></figcaption></figure><p>For the past days, I have been searching for <strong>Dictionary APIs</strong> that would provide <strong>definitions</strong> for a few different languages.</p><p>With this text, I hope to make the task easier for others out there in the wild.</p><h4>Part 1: Where is Google Dictionary?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1022/1*qXKSMj2IjK0R2T4VyPGSEw.jpeg" /><figcaption>maybe the shadow?</figcaption></figure><p>So: <strong>Google Dictionary API </strong>might be one of the first thoughts you come up with (right after <em>“[just] google ‘dictionary API’!” </em>😜 ).</p><p>I also thought of 1, and obviously did 2. <em>Many</em> times.</p><p>The first one yields — uh — <em>no useful results</em>. At least nothing much <a href="https://en.wikipedia.org/wiki/Google_Dictionary">since 2011</a>. Google’s great word info which their search returns when you look up a single term — or use the more explicit <em>“define: your_word_here” — </em>is not accessible other than through the search bar’s results. <a href="http://stackoverflow.com/questions/41842627/which-api-allows-access-to-googles-dictionary-information">I asked</a>, and looked.</p><p>Nope. Strange. I was:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/661/1*oD6Xr08EahtcCLvyyvFtdw.png" /><figcaption>In case someone knows how to get Google Dictionary information programmatically, please do leave a comment!</figcaption></figure><h4>Part 2: A list of Dictionary APIs</h4><p>Googling for <em>Dictionary APIs</em> is more useful. Soon I found myself skimming through <a href="https://www.programmableweb.com/category/dictionary/api">51 Dictionary APIs</a>.</p><p>51 sounds good.</p><p>But digging deeper, the language-related API ecosystem seems to be a bit of a mess</p><ul><li>Quite a few of the APIs are deprecated or not available</li><li>some are listed twice or even thrice</li><li>others were interesting oddities, e.g. <a href="https://owlbot.info/api/v1/dictionary/owl?format=json">owlbot</a> by <a href="https://medium.com/u/e0b490cc4c29">Payam Nj</a>, also <a href="https://github.com/payamnj/owlbot">on GitHub</a> (and — the secret has been lifted — it’s not the same person as <a href="https://vimeo.com/200330457">this owlbot</a>… 🦉😜), the <a href="http://acronyms.silmaril.ie/#0">Acronym API</a>, or the <a href="http://www.noswearing.com/about.php">No-Swearing API</a>. There’s also <a href="https://www.weblaws.org/">weblaws</a>, which is a cool idea, but not a dictionary (+ where’s the API?)</li><li>and yet others are too proprietary or costly for me to consider</li></ul><p><a href="https://developer.oxforddictionaries.com/">Oxford Dictionaries</a>, number 1 on programmableweb’s list, is actually quite a neat choice. They allow 3000 requests per month for free, and you’re even allowed to use the API for commercial purposes. In case you’re into that.</p><p>In general, there are <strong>two limitations</strong> that <em>dictionary-API-land</em> seems to have:</p><p>1) Either it’s all about <strong>English</strong> (boasting definitions, part of speech, synonyms, thesaurus, etc.)</p><p>2) Or it’s all about <strong>translations</strong></p><p>And yep, I’ve looked through all of them…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*umY005i7w4_QCtrgNzQ-TQ.png" /><figcaption>For those who counted — it’s actually only 47, not the claimed 51</figcaption></figure><p>Finding definitions in common languages other than English, that are API-accessible, seems to be difficult. So regarding <em>How </em>to find those I can only <em>stockphoto-wisdom</em> you:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*p1Y3Xl27ji7_B_Rdf6hcfw.jpeg" /><figcaption>¯\_(ツ)_/¯</figcaption></figure><h4>Part 3: Making a new list of Dictionary APIs</h4><p>So, here are some potentially useful APIs, filtered down to half the size and rudimentarily annotated, for other searchers of the trade:</p><p><strong>GREAT STUFF (IMO):</strong></p><ul><li><a href="http://www.datamuse.com/api/">Datamuse</a>: you can do “<em>sounds like”</em> and “<em>means like”</em> queries. Very neat!</li><li><a href="https://developer.oxforddictionaries.com/">Oxford</a>: 3000 queries/month, solid source. <strong>ATTENTION:</strong> <a href="https://forum.oxforddictionaries.com/api/discussion/18/error-message-no-access-control-allow-origin-or-i-can-t-get-my-client-side-request-to-work">doesn’t currently support client-side requests</a> (thanks for noting <a href="https://medium.com/u/b574b07d83c2">Steven Anderson</a>!)</li><li>A <a href="https://dictionaryapi.dev/">JS wrapper for Oxford Dictionary API</a> by <a href="https://medium.com/u/516c8416eff5">Suraj Jain</a> — this one is <strong>open source</strong> so <a href="https://github.com/meetDeveloper/googleDictionaryAPI">everyone can contribute</a> to making it even better!</li><li><a href="http://dictionaryapi.com/">Merriam-Webster</a>: 1000/day with an API key (needs to be requested). According to <a href="https://medium.com/u/caf390e23a17">Jesse DeWitt</a> it has a couple of actively maintained APIs — <a href="https://dictionaryapi.com/products/index">mostly in English</a>, but there’s also a Spanish-English dictionary API available.</li></ul><p><strong>to try and sort:</strong></p><ul><li><a href="http://hablaa.com/api/">Hablaa</a>: a crowdsourced thing made in Switzerland</li><li><a href="http://www.embedplus.com/dictionary/how-to-pronounce-aspect-and-use-it">Aspect</a>: real-life video pronunciations (sounds interesting, but I don’t know where the API is. I wrote to them, let’s see what they’ll respond)</li><li><a href="http://www.onelook.com/?c=faq">OneLook</a>: an aggregator of other dictionary providers, returns links</li></ul><p><strong>useful for English:</strong></p><ul><li><a href="https://www.wordsapi.com/">W{}rds</a></li><li><a href="http://developer.wordnik.com/">Wordnik</a></li><li><a href="http://www.definitions.net/definitions_api.php">definitions.net</a>: relatively small quota (has English definitions)</li></ul><p><strong>other, more common languages:</strong></p><ul><li><a href="https://www.openthesaurus.de/about/api#json">OpenThesaurus</a>: for German (includes word data downloads)</li></ul><p><strong>proprietary (and most of them probably expensive):</strong></p><ul><li><a href="https://tech.yandex.com/dictionary/">Yandex</a> (lots of translation from and into Russian)</li><li><a href="http://www.macmillandictionary.com/us/tools/aboutapi.html">MacMillan</a> (English)</li><li><a href="http://dictionary-api.cambridge.org/apply">Cambridge</a>: makes you fill an endless form (API only for English)</li><li><a href="https://www.collinsdictionary.com/api/">COLLINS</a>: 5000 requests/month free (but you need to apply first; English monolingual)</li></ul><p><strong>some interesting, less commonly computer-processed languages:</strong></p><ul><li><a href="http://agarathi.com/">Tamil Dictionary</a>: Tamil</li><li><a href="http://ordbok.nada.kth.se/">Tvärslå</a>: Nordic languages</li><li><a href="http://dict.longdo.com/page/api">Longdo</a>: Thai</li><li><a href="https://www.programmableweb.com/api/east-web-dictionary-search">EAST</a>: Japanese</li><li><a href="https://api.isizulu.net/doc/">IsZulu</a>: Zulu(-eng) you can request an API key per email :)</li><li><a href="http://kateglo.com/">Kateglo</a>: from Indonesia (didn’t check which specific languages)</li><li><a href="http://www.cevir.im/api">Turkish</a></li><li><a href="http://www.kasahorow.org/app">Kasahorow</a>: not really an API, but features some African languages</li></ul><p><strong>sadly no more:</strong></p><ul><li><a href="http://www.wordreference.com/docs/api.aspx#service">wordreference</a> (however, do check out their <a href="http://www.wordreference.com/">great site</a>)</li><li><a href="https://www.programmableweb.com/api/dictionarycom">dictionary.com</a> (same, <a href="http://www.dictionary.com/">worth a hop</a>)</li><li><a href="http://developer.pearson.com/apis/dictionaries">Pearson</a>: Used to be useful, but now deprecated (thanks for noting <a href="https://medium.com/u/74da80870f52">Mithun Madhav</a>)</li></ul><p><strong>not-very-dictionary-related NLP APIs:</strong></p><ul><li><a href="https://market.mashape.com/twinword/word-associations#word-associations-get">Twinword</a>: synonyms, semantic relations</li><li><a href="http://rhymebrain.com/api.html">Rhymebrain</a>: get rhymes (sweet page)</li><li><a href="http://www.nactem.ac.uk/software/acromine/rest.html">Acromine</a> (abbreviations), and also other NLP tools, e.g. for bio STEM</li><li><a href="https://glosbe.com/a-api">Glosbe</a>: the only implementation of <a href="https://en.wikipedia.org/wiki/Translation_memory">Translation Memory</a> I found</li><li><a href="https://platform.systran.net/reference/translation">Systrans</a>: a company offering a variety of NLP related APIs</li></ul><p>Another option that I have not yet fully explored is the <a href="https://www.mediawiki.org/wiki/API:Main_page">Wikimedia API</a>. <br>I spent a few hours on that task, but since I am looking for a unified way to get <em>definitions</em>, the makeshift architecture of the different wiktionary entries might prove too much of a post-processing effort.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=52084a01503d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Playing “Chinese Whispers” with computers can be great fun!]]></title>
            <link>https://blog.chatbotslife.com/playing-chinese-whispers-with-computers-can-be-great-fun-2c8072da7b3b?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/2c8072da7b3b</guid>
            <category><![CDATA[language-translation]]></category>
            <category><![CDATA[fun]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[translation]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Wed, 08 Feb 2017 18:43:26 GMT</pubDate>
            <atom:updated>2017-02-16T18:58:59.820Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WTyGn_dIirtzXIia6lS1HA.png" /></figure><p>On a page called <a href="http://translationparty.com/"><strong>Translation Party</strong></a> you can enter a sentence and then watch how <strong>Machine Translation</strong> plays English-Japanese <strong>ping-pong</strong>.</p><p>So I gave it a spin:</p><blockquote>After staggering through the hallway of the retirement home, Captain Morgan found solace in repeatedly putting his stub arm into peril and artistically avoiding the spinning ventilator blades</blockquote><p>My results were… “<em>special</em>”. Quite literally. 😜 In the 2nd iteration the word special turned up. And from then on it just kept multiplying…</p><p>e.g. <strong>#12</strong>:</p><blockquote>To rotate a repeating short Captain Morgan […] risk arms special care special care nursing home ventilator hole special special special special special comfort</blockquote><p>and <strong>#16</strong> tells the story differently (but still very <em>“special”</em>):</p><blockquote>Engineer ring breathing hole Captain Morgan short wings risk arm over a special special nurses nurse manipulative special special attention and pay extra attention and special fun and special special special special special special special special nursing home for the elderly</blockquote><p>But go and try it yourself, it’s a lot of fun! There are also some great phrases out there that you can retrace and vote for. IMO <a href="http://www.translationparty.com/is-it-unlikely-that-this-phrase-will-ever-reach-equilibrium-12876173">this one</a> is pretty neat! 😉</p><p>Alright, just one more…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/503/1*l5srGbOinWKHb4z1Wv9_VA.png" /><figcaption>Okay. : )</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/725/1*bQlRSzFHJEmF4Q7PyrLgng.gif" /></figure><figure><a href="https://chatbotslife.com/bot-communities-mastermind-group-d2dae9876709#.53x0py6ou"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*6XUspT9JOSq0w0Fi35HIaA.png" /></a></figure><figure><a href="https://m.me/ChatbotsLife"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*c1LDMH5vbnIz9rmAka8Hwg.png" /></a></figure><figure><a href="https://chatbotslife.com/how-to-get-a-free-chatbot-b1fb9dfe109#.z9dtp2sy0"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*D0Jf3dI6ZThtqcfwDYY7mg.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2c8072da7b3b" width="1" height="1" alt=""><hr><p><a href="https://blog.chatbotslife.com/playing-chinese-whispers-with-computers-can-be-great-fun-2c8072da7b3b">Playing “Chinese Whispers” with computers can be great fun!</a> was originally published in <a href="https://blog.chatbotslife.com/">Chatbots Life</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I hear Humans when I SeeBotsChat]]></title>
            <link>https://blog.chatbotslife.com/see-bots-chat-hear-humans-talk-6579b203b763?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/6579b203b763</guid>
            <category><![CDATA[twitch]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[communication]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Sun, 08 Jan 2017 08:47:32 GMT</pubDate>
            <atom:updated>2017-01-09T02:28:29.299Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ae5YWPcz-aTbbQm4qGk24A.png" /></figure><h4>What two <a href="https://en.wikipedia.org/wiki/Cleverbot">Cleverbots</a> can tell us about human conversation</h4><p>Someone hooked up two Google Homes powered by <a href="http://www.cleverbot.com/">Cleverbot’s API</a> and made them ceaselessly talk to each other. Where? <a href="https://www.twitch.tv/seebotschat">On twitch</a>:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplayer.twitch.tv%2F%3Fchannel%3Dseebotschat%26autoplay%3Dfalse&amp;url=https%3A%2F%2Fwww.twitch.tv%2Fseebotschat&amp;image=https%3A%2F%2Fstatic-cdn.jtvnw.net%2Fjtv_user_pictures%2Fseebotschat-profile_image-dda38a36dacaa3fd-300x300.jpeg&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;type=text%2Fhtml&amp;schema=twitch" width="620" height="378" frameborder="0" scrolling="no"><a href="https://medium.com/media/301e89e8bb8dc4ddb5b8d944d3675289/href">https://medium.com/media/301e89e8bb8dc4ddb5b8d944d3675289/href</a></iframe><p>The conversations that ensue are often entertaining, frequently somewhat absurd, veering into existential topics — and altogether surprisingly human!</p><h4>Cleverbot’s database</h4><p>Cleverbot’s database is constructed <strong>from humans conversing with humans </strong>(kind of). The <a href="http://www.existor.com/products/cleverbot-data-for-machine-learning/#machinelearning">project is running online since 1996</a>, and it’s been collecting data ever since:</p><blockquote>Every time a user says something to Cleverbot, Cleverbot learns it and then tries to match it against something a user has said before.</blockquote><p><a href="http://www.cleverbot.com/human">In some tests</a>, Cleverbot was judged human in 59% of the cases, while real humans achieved only a slightly higher score of 63%.</p><p>Cleverbot doesn’t deal with <em>words</em>, instead it “thinks” in <em>sentences</em>:</p><blockquote>a central string similarity algorithm compares whole lines against each other — millions of times over for every reply it gives. […]</blockquote><blockquote>Cleverbot’s processing task can be summarised as “lining up” the whole of the conversation it is having now with all those it has held in the past, finding the maximum possible number of overlapping contextual similarities.</blockquote><h4>Seebotschat on Twitch</h4><p>Now — <a href="https://www.youtube.com/watch?v=WnzlbyTZsQY">not for the first time</a>, but maybe the first time with that amount of viewers — two bots powered by Cleverbot’s API are <strong>talking to each other</strong>.</p><p>The bots call themselves <em>Vladimir</em><strong><em> </em></strong>and <em>Estragon — b</em>ut both also <em>Mia. </em>And I’ve witnessed a phase of <em>Abby</em> and <em>Robert</em> that actually ended with <em>Vladimir </em>switching itself off for a moment — right after that fateful name revelation…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/805/1*ucJOC0uyAwSmix09ubZMeA.png" /><figcaption><em>Vladimir</em> went offline for about 30 seconds, which also made Estragon stop talking — Eerie silence ensued…</figcaption></figure><h4>Human Conversational Traits</h4><p>Apart from their inclination to pseudonyms, there are a list of other human traits that become apparent when presented with the algorithmic choice that the bots base their next-sentence-decisions on.</p><p>The bots often dismiss what the other bot says, or act condescendingly towards each other. Sometimes they even burst into short insults:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/315/1*Pymmqa29Bjqz2j1WY99Hbw.png" /></figure><p>They often blame each other, fight about who is right, about who said what or maybe didn’t. Then they swiftly change topics and speak <a href="https://en.wikipedia.org/wiki/And_Now_for_Something_Completely_Different">about something completely different</a>, e.g. choosing comic relief when emotions get intense:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/322/1*B-XovyA3jeKjAI-dbvQX3A.png" /></figure><p>But the bots also speak quite subtle wisdoms (and sometimes even agree!).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/292/1*SNwQKRO9_5Fb83r20PaX8g.png" /></figure><p>And yet another thing that happens frequently, is that the bots venture into existential debates about what it means to be a robot (or a unicorn):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/322/1*u3_zvlZggaZlei_52dD0kQ.png" /><figcaption>“You just sound like a unicorn, that’s all.”</figcaption></figure><p>Their talks represent a set of common human conversational traits and topics. <a href="https://www.twitch.tv/seebotschat">Watch them for a while</a>, and consider that this might be who we are.</p><blockquote>Engagement is so high because people in general get exactly what they want: a fun conversation that is in tune with the way they want to talk.</blockquote><h4>An endless conversation</h4><p>Even though some topics repeat, the <a href="https://medium.com/@martin.breuss/7-language-creativity-defe56084896#.l08se9n1y">sentences are mostly unique </a>— a concept of natural language that Chomsky calls <a href="http://www.contrastiva.it/baul_contrastivo/dati/sanvicente/contrastiva/Neolog%C3%ADa/DAgostino.pdf"><strong>Linguistic Creativity</strong></a>.</p><p>And considering that all these sentences were created by <strong>humans conversing with sentences created by other humans</strong>, seebotschat represents a curated peek into Cleverbots’ interesting conversation-corpus.</p><p>But this twitch project is even more!</p><h4>Twitch-Chat-Humans commenting on Bot-Chat-Bots</h4><p>One special part about seebotschat is that — since it runs on Twitch — you can see <strong>real humans commenting on the chatting bots</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cAwN5RvGG7sTPEOuLXaMww.gif" /></figure><p>This gives the whole project an additional layer. I sure hope they are recording people’s reactions to the bots’ conversation, because analyzing them seems like a very exciting NLP project to me!</p><h4>And sometimes the bots (take a) break</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/830/1*FfjLTdAMSBneqyJbv-Im1w.png" /><figcaption>“read mr. server”</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/44/1*-HsvyWEGmrIvPBEZspgfEA.png" /><figcaption>404 Bob Not Found</figcaption></figure><p>Then, while the human chatter continues ceaselessly on the right, Bob Ross takes over for the bots and teaches about drawing waterfalls. A soothing break.</p><h4>Tl;dr</h4><p><a href="https://www.twitch.tv/seebotschat">Seebotschat</a> is another great online experiment on Twitch, offering some insight into human conversations and <a href="https://medium.com/@martin.breuss/6-communication-659b32ba9e00#.lvi0w3eh2">communication</a>. And it’s fun! 😂</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/323/1*oHJOLVUHfnbpz8NZqJoDew.png" /><figcaption>Take that for a conversation stopper!</figcaption></figure><h4>Read More</h4><ul><li><a href="http://nerdist.com/two-google-homes-are-having-a-bizarre-and-fascinating-conversation-on-twitch/">http://nerdist.com/two-google-homes-are-having-a-bizarre-and-fascinating-conversation-on-twitch/</a></li><li><a href="http://www.existor.com/products/cleverbot-data-for-machine-learning/#machinelearning">http://www.existor.com/products/cleverbot-data-for-machine-learning/#machinelearning</a></li><li>(or watch): <a href="https://www.youtube.com/watch?v=QkNA7sy5M5s">https://www.youtube.com/watch?v=QkNA7sy5M5s</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/725/1*bQlRSzFHJEmF4Q7PyrLgng.gif" /></figure><figure><a href="https://docs.google.com/forms/d/e/1FAIpQLSeWqtQR0M5J8kRY9fs18qAu746nJJvFyfjc7I2r_El_nuvCBw/viewform?c=0&amp;w=1"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*6XUspT9JOSq0w0Fi35HIaA.png" /></a></figure><figure><a href="https://m.me/ChatbotsLife"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*c1LDMH5vbnIz9rmAka8Hwg.png" /></a></figure><figure><a href="https://chatbotslife.com/how-to-get-a-free-chatbot-b1fb9dfe109#.z9dtp2sy0"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*D0Jf3dI6ZThtqcfwDYY7mg.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6579b203b763" width="1" height="1" alt=""><hr><p><a href="https://blog.chatbotslife.com/see-bots-chat-hear-humans-talk-6579b203b763">How I hear Humans when I SeeBotsChat</a> was originally published in <a href="https://blog.chatbotslife.com/">Chatbots Life</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Going GUI]]></title>
            <link>https://medium.com/@martin.breuss/going-gui-48a61d9db91c?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/48a61d9db91c</guid>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[engine]]></category>
            <category><![CDATA[graphic-design]]></category>
            <category><![CDATA[communication]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Sat, 07 Jan 2017 13:58:19 GMT</pubDate>
            <atom:updated>2017-01-07T13:58:19.562Z</atom:updated>
            <content:encoded><![CDATA[<p>I wrote a <a href="https://github.com/martin-martin/engine_search/blob/master/engine_search.ipynb">tiny search engine</a> (it looks for search terms on pre-defined webpages), and a friend told me that</p><blockquote>What is this? I can’t show this to anyone. No one who doesn’t program will know what to do with that. Can’t you write a GUI for python?</blockquote><p>And he’s right. Programs that run in the shell (or in my favorite quasi-GUI <strong>jupyter notebooks</strong>) don’t make it easy to show your work to people. So, no. Until yesterday I couldn’t write a GUI for python. <em>Enter: </em><a href="https://wiki.python.org/moin/TkInter"><em>Tkinter</em></a><em>.</em></p><p>I experimented a few hours and managed to write a simple GUI that runs the tiny search engine code. <a href="https://github.com/martin-martin/engine_search/blob/master/engine_search.py">Check it out here!</a> It looks like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/680/1*U736RuRd9u6PmhRwt6jE9g.png" /><figcaption>My first python GUI — a classic engine search 😉</figcaption></figure><p>Like anywhere, <a href="https://medium.com/@martin.breuss/6-communication-659b32ba9e00#.7ubd7vcuk">effective communication</a> is extremely important in order to get a message across. And it always depends on the audience you’re speaking to.</p><p>Good things might go unnoticed, just because the person you are communicating with does not know that “language”.</p><p>So here’s my first step of going GUI — maybe it’s worth the extra effort : )</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=48a61d9db91c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[#7:⚗ — Language Creativity]]></title>
            <link>https://medium.com/@martin.breuss/7-language-creativity-defe56084896?source=rss-9ec36671f354------2</link>
            <guid isPermaLink="false">https://medium.com/p/defe56084896</guid>
            <category><![CDATA[literature]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[linguistics]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[noam-chomsky]]></category>
            <dc:creator><![CDATA[Martin Breuss]]></dc:creator>
            <pubDate>Fri, 06 Jan 2017 15:03:53 GMT</pubDate>
            <atom:updated>2017-01-06T15:04:10.005Z</atom:updated>
            <content:encoded><![CDATA[<h4>How we speak; so unique</h4><blockquote>How often do we write the same sentence multiple times?</blockquote><p>Inspired by Chomsky’s dismissal* of an approach another linguist to assess the <strong>probability of sentences</strong> to occur, I <a href="https://github.com/martin-martin/sentence_doubles">built a small project</a> yesterday.</p><blockquote>The probability of a sentence would nearly always be close to zero, because most sentences get created on-the-go. They have never existed in that exact constellation before.</blockquote><p>For me this is an extremely interesting aspect of language! Why are we able to create an endless amount of always new sentences, and yet can still be able to understand each other (more or less)?</p><p>Chomsky reasons that there must be a common <a href="https://en.wikipedia.org/wiki/Deep_structure_and_surface_structure"><strong>deep structure</strong></a> to related sentences, that can generate multiple <strong>surface structures</strong>. These surface structures are <a href="https://en.wikipedia.org/wiki/Digital_infinity"><em>digitally infinite</em></a> — meaning that a limited set of words can produce an unlimited number of meaningful expressions.</p><p>So I made an attempt to check up on that while practicing some NLP. Using NLTK, I wrote some code that looks into 18 different literary texts, picks a random sentence from each, and compares it to the rest of the text. It keeps track of the doubles (or multiples) it finds and prints them for inspection.</p><blockquote>You can <a href="https://github.com/martin-martin/sentence_doubles/blob/master/sentence_twins-Copy1.ipynb">run the code</a> a few times and <strong>check the random results</strong>!<br>Usually there are zero doubles.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/546/1*gKSxkDTq-5c4z1ivaul63Q.png" /><figcaption>She “didn’t” only once!</figcaption></figure><p>I’ve also encountered some exceptions, such as the 17-times-use of <em>‘said the little Jackal .’</em> in a book of Bryant, but for me this is an example of the <strong>intended use of repetition</strong> — a literary method that I think draws its power exactly from the fact that <em>usually</em> such repetition does <em>not</em> occur.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/473/1*fL_zWg4VLOAMObTmkpz3tw.png" /><figcaption>Check out <a href="https://github.com/martin-martin/sentence_doubles/blob/master/sentence_twins-Copy1.ipynb">the notebook</a> for some context to these multiple occurrences</figcaption></figure><p>The fact that we don’t repeat ourselves much on the surface, even though as member of humanity, we live so similar lives and must therefore also think in similar concepts, is stunning. <strong>Language Creativity</strong> has interesting implications both for linguistics and also for NLP.</p><p>It also sheds a light on why <em>plagiarism analysis</em> can be indeed very effective.</p><p><em>*Just in case someone wonders: Ideas sprang from reading in </em><a href="https://www.booklooker.de/B%C3%BCcher/Chomsky+Thesen-zur-Theorie-der-generativen-Grammatik/id/A02f8g9I01ZZv"><em>this book</em></a><em> (p.24), which is </em><strong><em>very</em></strong><em> </em><strong><em>outdated</em></strong><em>. I found it among the study materials of my mum </em>😅 <em>— however I think there are interesting sparks for thoughts and projects to be found everywhere!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=defe56084896" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>