<?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[The Academy - Medium]]></title>
        <description><![CDATA[A collection of ideas, predictions, helpful hints, how-to’s, tools, and more… - Medium]]></description>
        <link>https://medium.com/the-academy?source=rss----7b9861ba4a01---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>The Academy - Medium</title>
            <link>https://medium.com/the-academy?source=rss----7b9861ba4a01---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 10:15:49 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/the-academy" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Exporting Data Out of Vercel KV]]></title>
            <link>https://medium.com/the-academy/exporting-data-out-of-vercel-kv-510f27d98afe?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/510f27d98afe</guid>
            <category><![CDATA[export]]></category>
            <category><![CDATA[vercel]]></category>
            <category><![CDATA[upstash]]></category>
            <category><![CDATA[redis]]></category>
            <category><![CDATA[backup]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Mon, 26 Jun 2023 20:06:19 GMT</pubDate>
            <atom:updated>2023-06-27T02:05:32.714Z</atom:updated>
            <content:encoded><![CDATA[<h3>TL;DR</h3><ol><li><a href="https://github.com/upstash/upstash-redis-dump/releases">Download</a>, extract, and prepare to run upstash-redis-dump</li><li>Run upstash-redis-dump -db 0 -host HOST -port PORT -pass PASSWORD -tls &gt; redis.dump</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sByE7_KkQjIJimXji2WX5Q.png" /></figure><h3>More Info</h3><p>Vercel <a href="https://vercel.com/docs/storage">offers</a> easy to setup storage products. One such product is called <a href="https://vercel.com/docs/storage/vercel-kv">Vercel KV</a>, which is powered by Redis through a partnership with <a href="https://upstash.com/">Upstash</a>. However, it doesn’t include all of the Redis-native features, as seen in <a href="https://vercel.com/docs/storage/vercel-kv/limits">this list</a> of unsupported features. The primary one being the ability to backup and restore the data.</p><p>I received an email from Vercel saying that I’ve run out of 30,000 requests. <a href="https://cdn-images-1.medium.com/v2/resize:fit:1600/1*Qm0JmNTeG15NVQUFCEEE_A.png">That wasn’t even true.</a> But by that point, I already learned of Upstash’s free tier, which is much more generous than Vercel’s “Hobby” tier:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/639/1*We0qFQMeLZ2MqkIIkf4Plw.png" /><figcaption>(Source: <a href="https://wundergraph.com/blog/dodging-the-vercel-tax">WunderGraph — Dodging the Vercel Tax</a>)</figcaption></figure><p>All together, that prompted me to try and back up all of my data so I could migrate it someplace else. I ended up spending a bunch of time attempting to export it out. At first via <a href="https://redis.io/docs/getting-started/">the </a><a href="https://redis.io/docs/getting-started/">redis-cli</a> as recommended by the docs:</p><pre>$ redis-cli --tls -u redis://username:password@subdomain.domain.tld:54321 --rdb redis.rdb<br>Warning: Using a password with &#39;-a&#39; or &#39;-u&#39; option on the command line interface may not be safe.<br>sending REPLCONF capa eof<br>REPLCONF capa error: ERR Command is not available: &#39;REPLCONF&#39;. See https://vercel.com/docs/errors/vercel-kv-error-codes for details<br>sending REPLCONF rdb-only 1<br>REPLCONF rdb-only error: ERR Command is not available: &#39;REPLCONF&#39;. See https://vercel.com/docs/errors/vercel-kv-error-codes for details<br>SYNC with master failed: -ERR Command is not available: &#39;SYNC&#39;. See https://vercel.com/docs/errors/vercel-kv-error-codes for details</pre><p>When that failed, I tried to run the <a href="https://redis.io/commands/save/">SAVE</a> command right from the console:</p><pre>$ redis-cli --tls -u redis://username:password@subdomain.domain.tld:54321<br>Warning: Using a password with &#39;-a&#39; or &#39;-u&#39; option on the command line interface may not be safe.<br>subdomain.domain.tld:54321&gt; SAVE<br>(error) ERR Command is not available: &#39;SAVE&#39;. See https://vercel.com/docs/errors/vercel-kv-error-codes for details</pre><p>…but it also failed. As did <a href="https://redis.io/commands/bgsave/">BGSAVE</a>.</p><p>I then looked into writing a script to pull the data out, key-by-key. First, I explored the option of following <a href="https://github.com/RedisLabs/redis-migrate">existing scripts</a> that were mostly written in <a href="https://github.com/mingrammer/redis-migrate">Python</a> and <a href="https://github.com/obukhov/go-redis-migrate">Go</a>. That didn’t work, likely due to <a href="https://vercel.com/docs/storage/vercel-kv/limits#tls-support">Vercel KV’s TLS requirement</a>:</p><blockquote>When you create a Vercel KV store, a Redis store is provisioned with support for TLS v1.2 and v1.3 enabled. There is no way to disable TLS for your KV store.</blockquote><p>Lastly, when I tried to use <a href="https://www.npmjs.com/package/@vercel/kv">the </a><a href="https://www.npmjs.com/package/@vercel/kv">@vercel/kv library</a> outside of the Next.js framework, it kept throwing a “connection refused” error. It’s worth mentioning that Vercel also created Next.js. And just as I was about to give up, I came across this “<a href="https://docs.upstash.com/redis/howto/importexport">Import/Export Data</a>” document from Upstash. That’s where I learned about their nifty tool called <a href="https://github.com/upstash/upstash-redis-dump">upstash-redis-dump</a>.</p><h3>Using upstash-redis-dump</h3><p>Below you will find more detailed instructions on how to export and import your data out of Vercel KV and into either a self-hosted Redis instance or onto Upstash.</p><h4>Prerequisites</h4><ol><li>Download the appropriate upstash-redis-dump <a href="https://github.com/upstash/upstash-redis-dump/releases">release</a><br>(E.g., <a href="https://github.com/upstash/upstash-redis-dump/releases/download/v0.1.0/upstash-redis-dump_macos_arm64.zip">upstash-redis-dump_macos_arm64.zip</a> that runs on Apple’s chips)</li><li>Extract the binary out of the zip file</li><li>Turn it into an executable by running chmod +x on it in the Terminal</li></ol><h4>Export</h4><p>Export the data using the following command:</p><pre>upstash-redis-dump -db 0 -host HOST -port PORT -pass PASSWORD -tls &gt; redis.dump</pre><p>Be sure to replace HOST, PORT, and PASSWORD using the settings from <a href="https://vercel.com/dashboard/stores">your Vercel Storage dashboard</a>.</p><h4>Import</h4><p>Import to Upstash or any other Redis instance:</p><pre>redis-cli --tls -u redis://username:password@subdomain.domain.tld:54321 --pipe &lt; redis.dump</pre><p>You’ll want to replace pretty much everything in that URL. Just be mindful of the --tls flag, which may not be applicable in your case.</p><h3>Closing Thoughts</h3><p>I sincerely hope you found this helpful. Please share any feedback by commenting below. And let me know if you know of any other solutions for rapid prototyping whilst minimizing vendor lock-in!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=510f27d98afe" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/exporting-data-out-of-vercel-kv-510f27d98afe">Exporting Data Out of Vercel KV</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Deploying AI to the Cloud]]></title>
            <link>https://medium.com/the-academy/deploying-ai-to-the-cloud-3759d04df679?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/3759d04df679</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[banana]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[serverless]]></category>
            <category><![CDATA[text-to-speech]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Wed, 11 Jan 2023 18:21:44 GMT</pubDate>
            <atom:updated>2023-01-12T16:21:01.021Z</atom:updated>
            <content:encoded><![CDATA[<p>This is a tale of the hurdles I had to overcome in my pursuit of deploying an artificial intelligence based program to the cloud for general use. But first, a quick disclaimer and a TL;DR.</p><p><strong>Disclaimer:</strong> while I used AI in the past — voice assistant, code completion, image generation — I’m new to the concept of AI development. So, if you come across any glaring errors or obvious shortcomings, kindly share your thoughts and suggestions in a response below.</p><p><strong>TL;DR</strong>: in my experiments with high-quality text-to-speech conversions, I found that generating the voice samples locally was <em>sixty times</em> faster than doing so in an environment for GPU-enabled serverless computing. Therefore, I don’t recommend going after the same outcome at this time. Instead, consider using this article as a guide of what to expect when you’re ready to deploy your own AI creation to the cloud.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SyeojnkNqPV_Csqrq3h3gw.png" /><figcaption>“A tortoise on a flat road looking far into the distance. 4K 8K HDR photorealistic” – generated with Stable Diffusion v1.5 via <a href="https://dreamstudio.ai/">DreamStudio Beta</a></figcaption></figure><h3>Let’s Start With a Few Basic Concepts</h3><ul><li>Serverless Framework — an architecture that allows developers to deploy code to the cloud in a way that can be consumed by any other app (mobile, desktop, web, etc.) while only being charged for its runtime, which is usually measured in seconds. For this post, I’ll be using <a href="https://banana.dev/">Banana</a>, which currently costs roughly <em>one-twentieth of a US penny</em> ($0.00052) per one second of GPU use.</li><li>Local Testing — one or more physical machines that are dedicated to debugging, running, testing, and validating the application before it is published. Thus saving time and money by catching incomplete models and buggy code before it is deployed.</li><li>Docker — a virtualization platform that enables developers to deploy their code in a portable manner so that it can run on all kinds of servers with different configurations like CPUs, GPUs, and RAM, thus allowing one to easily scale the app as needed.</li><li><a href="https://github.com/neonbjb/tortoise-tts">Tortoise TTS</a> — an open source text-to-speech (“TTS”) program that can run on Windows OS or in a Google Collab notebook. It comes with more than twenty pre-trained, high-quality voices in terms of prose and intonation, including the ability to train infinitely more voices. Tortoise can also understand prompts such as <em>[I am really sad]</em> and use that information to modify the tone of voice accordingly (<a href="https://github.com/neonbjb/tortoise-tts#prompt-engineering">source</a>).</li></ul><h3>Implementation</h3><p><a href="https://docs.banana.dev/banana-docs/core-concepts/inference-server/serverless-framework">Banana’s Serverless Framework</a> is pretty straightforward. It offers a REST API, as well as SDKs in Python, Go, and Node that enable the developers to communicate with their creations. The call to Banana can include a collection of model inputs. Those are used for things like parameters, prompts, and properties that are used to produce and fine tune the models’ outputs.</p><p>In my case, I wanted to start by providing it with just the text that it needs to convert to audio. But per Tortoise’s own API, it can be expanded to accept inputs such as voices and their presets (existing or newly-trained), as well as multiple parameters that adjust the final output.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IpbVwf7RNGFH0oLz3mJTZA.png" /><figcaption>(<a href="https://github.com/neonbjb/tortoise-tts/blob/217dc09d5f81a45778d61c875ad6eaf70f6c7a92/tortoise/do_tts.py#L10-L25">source</a>)</figcaption></figure><p>Because the output is an audio file, there are basically two ways to play it back. The first way follows some examples of the existing text-to-image generative AI programs, where it’s possible to encode the output as a base-64 string. But instead of an image, it’s encoding an audio sample. This file can be then played back using the &lt;audio controls autoplay data=&quot;data:audio/wav;base64,Base+64+string+goes+here&quot; /&gt; HTML element. The other way is by storing the generated audio file on some server (e.g., AWS S3, SoundCloud, Google Drive, etc.) and returning a path to its location. Due to the experimental nature of this project, I chose to return the response as a base-64 string, mainly because I did not want to add to the complexity of hosting the resulting files elsewhere in the cloud. At this stage, it’s important to point out that one of the shortcomings of the Serverless Framework is that it’s mostly ephemeral. Meaning that with the common exception of the logs that it produces in real time, it leaves behind very few artifacts (if any) which are normally used in other types of software development projects.</p><p>The next challenge that I faced was downloading all of the pre-trained models from Hugging Face. They come out to 5.5GB. While Banana <a href="https://docs.banana.dev/banana-docs/core-concepts/inference-server/serverless-framework/download.py">offers</a> a convenient download.py script for this exact reason, I noticed that by taking this path, my development efforts were considerably slower due to the Docker containers needing to download these models every time they were created. So instead, I put together a new Docker image called <a href="https://hub.docker.com/repository/docker/ilyakam/deps-for-tts/general">deps-for-tts</a> that’s based on <a href="https://hub.docker.com/layers/pytorch/pytorch/1.11.0-cuda11.3-cudnn8-runtime/images/sha256-9904a7e081eaca29e3ee46afac87f2879676dd3bf7b5e9b8450454d84e074ef0?context=explore">PyTorch’s CUDA 11.3 image</a> that I borrowed from Banana’s existing <a href="https://docs.banana.dev/banana-docs/resources/list-of-banana-model-repos#banana-1-click-deploy-model-source-code">model repos</a>. All of the models were cached within it.</p><p>With the image and framework in tow, it was a relatively straightforward matter of creating a server and exposing a couple of predefined routes. One GET /healthcheck to ensure that everything’s running smoothly and that the GPU is ready. And another POST / that can take the model inputs and pass them to the application. Given that my code is open source, I see no reason of going into the technical details of its implementation. Instead, I encourage you to take a look at the following files in order to see how everything comes together:</p><ul><li><a href="https://github.com/ilyakam/tortoise-tts/blob/main/server.py">server.py</a> — the aforementioned REST server and route handler that follows the Serverless Framework best practices</li><li><a href="https://github.com/ilyakam/tortoise-tts/blob/main/app.py">app.py</a> — the application handler that calls the AI program, waits for its output, and encodes it for consumption on the frontend</li><li><a href="https://github.com/ilyakam/tortoise-tts/blob/main/tortoise/api.py">tortoise/api.py</a> — (unmodified) the entrypoint to Tortoise</li></ul><h3>Results</h3><p>Below is a comparison of the logs between my local development machine and what took place on Banana. My testing was done on an i5–3330 CPU with 16GB of DDR3 RAM and the main component being an Nvidia GeForce RTX 3060 video card that I bought specifically for this purpose.</p><h4>Local</h4><pre>Generating autoregressive samples..<br>100%|█████████| 12/12 [01:04&lt;00:00,  5.41s/it]<br>Computing best candidates using CLVP<br>100%|█████████| 12/12 [00:05&lt;00:00,  2.08it/s]<br>Transforming autoregressive outputs into audio..<br>100%|█████████| 80/80 [00:12&lt;00:00,  6.56it/s]</pre><h4>Banana</h4><pre>Generating autoregressive samples..<br> 33%|███▎      | 32/96 [2:26:21&amp;lt;5:20:02, 300.04s/it]<br> 67%|██████▋   | 64/96 [5:05:15&amp;lt;2:48:35, 316.11s/it]<br>100%|██████████| 96/96 [7:31:45&amp;lt;00:00, 282.35s/it]<br>Computing best candidates using CLVP<br><br>Transforming autoregressive outputs into audio..</pre><h4>Of note…</h4><ol><li>Compare the number of seconds per iteration (<em>n</em>s/it); fewer is better</li><li>Not output was provided on Banana for the second and third steps of computing via CLVP and transforming the autoregressive outputs</li><li>Instead of showing 0–100%, I truncated Banana’s output for brevity</li><li>7½ hours of GPU time at 0.05¢ per second comes out to about $14</li></ol><p>I’m not comfortable with including the actual audio samples that Tortoise had generated since I didn’t do anything to improve or modify the underlying architecture of that project. But if you’re interested in listening to what Tortoise produces, check out the project’s own <a href="https://github.com/neonbjb/tortoise-tts/tree/main/examples">examples folder</a>.</p><h3>Next Steps</h3><p>Now that I have an understanding how to deploy AI-powered programs to the cloud, I want to focus on developing and optimizing the machine learning algorithms themselves. My plan is to go through the “<a href="https://www.youtube.com/playlist?list=PLfYUBJiXbdtSvpQjSnJJ_PmDQB_VyT5iU">Practical Deep Learning for Coders 2022</a>” course. Consider following <a href="https://academy.ilyakam.com/"><em>The Academy</em></a> for more insights on this topic.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3759d04df679" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/deploying-ai-to-the-cloud-3759d04df679">Deploying AI to the Cloud</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Samsung Galaxy Z Flip3 is Not the Right Smartphone for Me]]></title>
            <link>https://medium.com/the-academy/the-samsung-galaxy-z-flip3-is-not-the-right-smartphone-for-me-ce0b07b30d8d?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/ce0b07b30d8d</guid>
            <category><![CDATA[comparison]]></category>
            <category><![CDATA[samsunggalaxyzflip3]]></category>
            <category><![CDATA[android]]></category>
            <category><![CDATA[review]]></category>
            <category><![CDATA[smartphones]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Tue, 14 Sep 2021 17:15:42 GMT</pubDate>
            <atom:updated>2021-09-14T18:15:10.678Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CwTp7Anwl-l8-t3zeL8CHQ.jpeg" /><figcaption>Samsung Galaxy S :: Huawei Nexus 6P :: Essential PH-1 :: Samsung Galaxy S10e :: Samsung Galaxy Z Flip3 5G :: Ruler</figcaption></figure><p>The largest smartphone I have ever owned was the Huawei Nexus 6P that was sold by Google. The device was massive at 159.3 × 77.8 mm (6.27 × 3.06 in) which made one-handed operation practically impossible. I made the decision to buy it after having the OnePlus One for about two years prior. That phone was also hefty at 152.9 ⨉ 75.9 mm (6 ⨉ 2 in). At the time I thought, “it’s just a marginal increase in size and it won’t make a difference.” But after struggling with both of those devices for about four years, I made the decision to only buy smaller smartphones that fit it one hand and can be used without having to slide them up and down.</p><p>This took me on a journey to first try the Blackberry KeyONE and the Samsung Galaxy S8 with <a href="https://www.samsung.com/us/mobile/mobile-accessories/phones/galaxy-s8-keyboard-cover--black-ej-cg950bbegww/">the Keyboard Case</a> shortly thereafter. Both of those phones had clocked in at about 149 ⨉ 70 mm (5.87 ⨉ 2.76 in). I was unimpressed with the processor and the camera on the Blackberry, so I exchanged it for the Samsung S8. While vacationing in Paris shortly after getting the S8, I vividly remember riding the metro to the Arc de Triomph and double-checking the station where I needed to get off on that device. After getting out of the train, there were about three flights of stairs, and by the time I got to the surface level and looked at my new phone once again, I noticed a hairline crack across the top of the screen even though it was kept in the original Samsung case and in my pocket the whole time between the train and the monument. So I returned it for being defective. Then I settled on the Essential PH-1. Despite having an ugly notch and no headphone jack, the phone was smaller (141.5 ⨉ 71.1 mm or 5.57 ⨉ 2.80 in), made up of uncommon materials (ceramic and titanium) that were great to the touch, and came with the best internals at the time.</p><p>The biggest downside to the PH-1 was its capacitive touchscreen. It was difficult to use, especially in light of the Android 10 Gesture Navigation where it would constantly miss the edge swipes to go back. So I ended up buying a Samsung Galaxy S10e with a similar size of 142.2 ⨉ 69.9 mm (5.60 ⨉ 2.75 in). Besides having an upgraded processor and an impeccable screen, the new device brought back old features such as the headphone jack, memory expansion via a microSD card, and IP68 water and dust resistance. Now, two years later, it’s starting to show its age. It’s the usual stuff; Samsung is about to stop supporting it right after <a href="https://www.androidcentral.com/samsung-one-ui-40-android-12">upgrading it to Android 12</a>, there’s no 5G, and the battery doesn’t last as long as it used to. (Though AccuBattery is saying that the battery’s health is at 95%, so it might be fixed with a factory reset.)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uhjB6MV9AqcDAftoQGaWUA.jpeg" /></figure><p>Which finally brings us to the $1000 Samsung Galaxy Z Flip3 5G. It is such a disappointment that I’m having a hard time deciding where to start. I guess I’ll start with the good — that Green color is absolutely fantastic! The rest of it simply sucks. Let’s look at the heft. It’s about as heavy as the PH-1 at 185 grams (6.5 oz) but it’s not as well-balanced. Samsung got a lot of criticism for its predecessors not being able to remain half open on the desk, so they made it bottom-heavy. Which means it’s less optimal for hand-held operation. Then there’s the sheer size itself. When unfolded, it’s more than 6mm taller than the tallest phone I’ve ever owned (166mm vs. the 6P’s 159.3mm). It is so tall and unbalanced that even when using Samsung’s one-handed feature, it’s still impossible to reach the top of the shrunken virtual screen:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n9lwRkk5eXLqJJ7TgWAjdA.jpeg" /><figcaption>Samsung Galaxy S10e :: Samsung Galaxy Z Flip3</figcaption></figure><p>The next big issue I have is with the way the screen feels. Even my very first Android phone, the Samsung Galaxy S (“Vibrant”) had Corning Gorilla Glass on the front. The Z Flip3’s main screen is covered with plastic. PLASTIC! This makes it feel almost like the <a href="https://www.gsmarena.com/samsung_t746_impact-2870.php">Samsung Highlight</a>. “Almost” because the feature phone didn’t have a noticeable dip halfway down its screen.</p><p>The rest are just stupid omissions that are inexcusable at this price point as compared to my current S10e (which I bought “renewed” for $385):<br>- No memory expansion<br>- No dedicated assistant button<br>- Out-of-reach volume buttons<br>- Unimpressive cameras<br>- And no headphone jack (which I still use about once every fifty days)</p><p>And all of that is traded for the fact that you can fold it and carry it your pocket, never mind that it bulges out and weighs you down like two halves of a chunky smartphones that was cut in half and stacked one atop the other.</p><p>Where’s my Android equivalent of the iPhone 13 Mini that comes with a 5G capable processor and a full-day battery?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ce0b07b30d8d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/the-samsung-galaxy-z-flip3-is-not-the-right-smartphone-for-me-ce0b07b30d8d">The Samsung Galaxy Z Flip3 is Not the Right Smartphone for Me</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Unconfirmed. We Live in a Simulation?]]></title>
            <link>https://medium.com/the-academy/unconfirmed-we-live-in-a-simulation-da60e21e01fe?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/da60e21e01fe</guid>
            <category><![CDATA[cpu]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[proof]]></category>
            <category><![CDATA[rebuttal]]></category>
            <category><![CDATA[simulation-hypothesis]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Tue, 06 Apr 2021 19:07:50 GMT</pubDate>
            <atom:updated>2021-04-06T19:07:50.225Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vMJmNsaJuOqLUkB0QsaPbw.jpeg" /></figure><p>This article is a short rebuttal to Fouad Khan’s piece titled “Confirmed! We Live in a Simulation” that was published in <a href="https://www.scientificamerican.com/article/confirmed-we-live-in-a-simulation/">Scientific American</a> on April 1, 2021.</p><p>In that piece, the author argues that there’s a theoretical maximum limit to the processor speed. And the entities that live in a simulation are aware of it.</p><blockquote>All computing hardware leaves an artifact of its existence within the world of the simulation it is running. This artifact is the processor speed. If for a moment we imagine that we are a software program running on a computing machine, the only and inevitable artifact of the hardware supporting us, within our world, would be the processor speed.</blockquote><p>Recall that in order to measure the velocity of a moving object, you must solve the equation of distance traveled over time. Similarly, a processor’s speed is measured by the amount of calculations in a given unit of time, where 1GHz roughly translates to one billion calculations per second. Given that speed is a function of time, let’s see how <a href="https://simple.wikipedia.org/wiki/Space-time">the Simple English Wikipedia article</a> talks about it in the context of “spacetime.”</p><blockquote>[…] in a relativistic universe, time cannot be separated from the three dimensions of space.</blockquote><p>If we were to create a simulation of our reality from scratch, we must include both concepts of time and space and apply them to everything therein. As such, time becomes a part of the simulation, and it too is therefore governed by the limits that are imposed by the processor.</p><p>When put together, this means that no entity within the simulation can reliably determine the speed at which the simulation is running. Sure, it can hit measurable limits like the speed of light (<strong><em>c</em></strong>) or Planck’s constant (<strong><em>h</em></strong>); however, they can only be measured against a relative concept of time as seen within the simulation itself. Thus, without an external anchor outside of the simulation, these constant values are not enough to prove or disprove that we live in one.</p><p>Anyone who has basic knowledge of computer programming is welcome to prove me wrong by solving the following challenge. I’ve set up a simulation boilerplate in JavaScript. The goal is to figure out the value of the variable called 𝚜𝚒𝚖𝚞𝚕𝚊𝚝𝚘𝚛_𝚌𝚙𝚞_𝚍𝚎𝚕𝚊𝚢_𝚒𝚗_𝚖𝚒𝚕𝚕𝚒𝚜𝚎𝚌𝚘𝚗𝚍𝚜 from within the function called 𝚛𝚞𝚗_𝚜𝚒𝚖𝚞𝚕𝚊𝚝𝚒𝚘𝚗 without using global objects like 𝙳𝚊𝚝𝚎 or 𝚌𝚘𝚗𝚜𝚘𝚕𝚎.𝚝𝚒𝚖𝚎.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d7b6d9087c586a006d668e2432bd12b8/href">https://medium.com/media/d7b6d9087c586a006d668e2432bd12b8/href</a></iframe><p>To be clear, I’m not saying that we don’t live in a simulation. I just haven’t come across any compelling proof that we do. I often think <a href="https://twitter.com/search?q=(planck%20OR%20&quot;speed%20of%20light&quot;%20OR%20simulation)%20(from%3Ailyakam)">and tweet</a> about this very problem. Follow me <a href="https://twitter.com/ilyakam">on Twitter</a> or <a href="https://academy.ilyakam.com/">on Medium</a> for more on this topic.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=da60e21e01fe" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/unconfirmed-we-live-in-a-simulation-da60e21e01fe">Unconfirmed. We Live in a Simulation?</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Control React Dropzone Uploader with React Hook Form]]></title>
            <link>https://medium.com/the-academy/control-react-dropzone-uploader-with-react-hook-form-99e8419546d6?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/99e8419546d6</guid>
            <category><![CDATA[react-hook-form]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[react]]></category>
            <category><![CDATA[dropzone]]></category>
            <category><![CDATA[source-code]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Wed, 18 Nov 2020 08:17:48 GMT</pubDate>
            <atom:updated>2020-11-18T08:17:48.153Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2PdLLQr46s8TEhLCWEkk-A.png" /></figure><p>From the rubric of “write the blog post that you wish already existed,” I wanted to share a straightforward way of wrapping the <a href="https://react-dropzone-uploader.js.org/">React Dropzone Uploader</a> component in a <a href="https://www.react-hook-form.com/">React Hook Form</a> controller. And in the process, to explore each of these popular libraries a bit more in-depth. Let’s get straight to the solution by starting with the full CodeSandbox here:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fcodesandbox.io%2Fembed%2Ftfvui&amp;display_name=CodeSandbox&amp;url=https%3A%2F%2Fcodesandbox.io%2Fs%2Ftfvui&amp;image=https%3A%2F%2Fcodesandbox.io%2Fapi%2Fv1%2Fsandboxes%2Ftfvui%2Fscreenshot.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=codesandbox" width="1000" height="500" frameborder="0" scrolling="no"><a href="https://medium.com/media/bdca4a6e17006be149a0d0eaef57a04a/href">https://medium.com/media/bdca4a6e17006be149a0d0eaef57a04a/href</a></iframe><h4>Controller</h4><p>React Hook Form claims to “embrace uncontrolled components.” That’s a bold statement. It hinges on the powerful &lt;Controller /&gt; component and the API that it provides. As always, I recommend <a href="https://react-hook-form.com/api#Controller">starting with the documentation</a>. In this case, we need to pay attention to the render method and its properties:</p><blockquote>onChange, onBlur, name, ref, <em>and</em> value</blockquote><p>onChange is a bit of a misnomer. I presume that it’s called onChange so you could easily write onChange={(data) =&gt; onChange(data)}. It also nicely shorthands to just onChange={onChange}.</p><p>But it’s more of a setter. In the sense that you call it with the value that you want to pass from the controlled input. So if the user types something into an input field, you’d want to onChange that value in order to set it to the named form data object. A more concrete example is when it’s combined with other handlers, like you can see on line 50 of src/App.tsx. I’ve renamed it to setFiles in the handler per my remark above. And that callback is called when the status of the files changes. Which brings us to…</p><h4>setTimeout</h4><p>At the time of this writing, <a href="https://github.com/fortana-co/react-dropzone-uploader/issues/126">there is a bug</a> in React Dropzone Uploader where the callback handleChangeStatus is called before the array of files is updated once a file is removed. A dirty fix to get around that is by using a no delay setTimeout that takes an action on the next tick of the <a href="https://reactjs.org/docs/state-and-lifecycle.html">React lifecycle loop</a>. When combined together with the React Hook Form controller, it correctly reacts to changes to the files array, even when a file is removed.</p><h4>Questions? Comments? Errors?</h4><p>Feel free to reach out below or <a href="https://twitter.com/ilyakam">on Twitter</a> if there’s anything I need to address.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=99e8419546d6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/control-react-dropzone-uploader-with-react-hook-form-99e8419546d6">Control React Dropzone Uploader with React Hook Form</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Serverless: How to Publish Your App and Not Your Source Code]]></title>
            <link>https://medium.com/the-academy/serverless-how-to-publish-your-app-and-not-your-source-code-743e884d22d6?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/743e884d22d6</guid>
            <category><![CDATA[hacking]]></category>
            <category><![CDATA[source-code]]></category>
            <category><![CDATA[serverless]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Sun, 10 Nov 2019 22:32:22 GMT</pubDate>
            <atom:updated>2019-11-10T22:32:20.938Z</atom:updated>
            <content:encoded><![CDATA[<p>Serverless platforms are great. Some of them allow you to publish apps without having to pay a dime. The caveat is that your source code must be public as well. That’s perfectly fine for small experiments and open source projects, but what if you want to keep your code to yourself? Well, with the power of two-way encryption, private environment variables, and an eval() function, it is possible to keep your apps public and your code private. Continue reading below to find out how.</p><p>For this demonstration, we’re going to look at the “Obfuscated Source Code Experiment.” I’ve set it up on RunKit, which is a Serverless platform for Node.js. You can see the final version of the app by navigating over to the published endpoint at <a href="https://obfuscated-source-code-experiment-4rfhxf9cxpqi.runkit.sh/">https://obfuscated-source-code-experiment-4rfhxf9cxpqi.runkit.sh</a>. But, when you look at <a href="https://runkit.com/ilyakam/obfuscated-source-code-experiment">the source code</a>, this is all you see:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c44d2ce60fa945d9135c25d48634fcbe/href">https://medium.com/media/c44d2ce60fa945d9135c25d48634fcbe/href</a></iframe><p>In the meantime, the actual source code for the application can be safely stored locally or in a private repository. Though you wouldn’t normally see it, here’s what it looks like for the app above:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/08d6b763d26559fe47666fa458a52531/href">https://medium.com/media/08d6b763d26559fe47666fa458a52531/href</a></iframe><p>And here is the boilerplate code to encrypt the secretFunction into a string and to print it in the console:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d43970e33241210aa7cbfc061debdbee/href">https://medium.com/media/d43970e33241210aa7cbfc061debdbee/href</a></iframe><p>All of this is enabled by the magic of environment variables. They are an integral part of the Serverless paradigm, allowing the developers to keep their secret values private. They’re generally used for things like authentication tokens, access keys, and configuration variables. You can learn a lot more about them by reading <a href="https://12factor.net/config">the third rule of the Twelve-Factor App guide</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1E_GkEoEmoql22KdicuYnQ.png" /><figcaption>You can set secret environment variables <a href="https://runkit.com/settings/environment">on RunKit</a></figcaption></figure><p>Disclaimer: I am not a security expert. Do your own research to find a secure package that offers two-way encryption that suits your needs. I only picked <a href="https://www.npmjs.com/package/simple-crypto-js">simple-crypto-js</a> because it worked well enough for this proof-of-concept.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=743e884d22d6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/serverless-how-to-publish-your-app-and-not-your-source-code-743e884d22d6">Serverless: How to Publish Your App and Not Your Source Code</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Actually Making Money by Trading Cryptocoins]]></title>
            <link>https://medium.com/the-academy/actually-making-money-by-trading-cryptocoins-aeb35c47fe6f?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/aeb35c47fe6f</guid>
            <category><![CDATA[ethereum]]></category>
            <category><![CDATA[cryptocurrency]]></category>
            <category><![CDATA[bitcoin]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Wed, 27 Sep 2017 20:58:07 GMT</pubDate>
            <atom:updated>2017-09-27T20:58:07.154Z</atom:updated>
            <content:encoded><![CDATA[<p>I’m sure you’ve heard of people making money by trading cryptocoins. The issue is that many of them are becoming rich “on paper.” Meaning, they don’t see a real dollar from their portfolio because the money is tied up in the exchanges. I wanted to devise an experiment where the only thing that counts is the bottom line. That is, the difference between the money that leaves my checking account and the amount that is coming back. The other objectives were to learn to trade multiple coins on different exchanges and in different ways.</p><p>Let’s start with the results first:</p><ul><li><strong>Ethereum:</strong> gain of $71.71 (+23.90%) over eleven days</li><li><strong>Neo:</strong> gain of $125.94 (+41.98%) over eleven days</li><li><strong>Bitcoin:</strong> gain of $109.98 (+18.33%) over thirteen days</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/815/1*PJL3j3nXyMeamyx4hIBgtA.png" /><figcaption>The other $300 came from my credit card because I wanted to buy Neo right away.</figcaption></figure><h4>Ethereum — the Easy Way</h4><ol><li>Purchase Ethereum on Coinbase by heading over to the Buy/Sell page, selecting Buy, choosing Ethereum, using my previously-verified checking account as a payment method, entering $300, and pressing the large Buy button. The USD⨉ETH rate is locked at this point.</li><li>Wait for the price to go up.</li><li>Sell Ethereum on Coinbase by heading over to the same Buy/Sell page, selecting Selling, choosing the “ETH Wallet,” depositing back to my checking account, typing the amount of Ethereum bought in the first step, and pressing the large Sell button.</li></ol><h4>Neo (and other fringe cryptocoins)</h4><ol><li>Purchase $300 worth of Bitcoin on Coinbase similarly to Ethereum’s first step above.</li><li>Send Bitcoin over to <a href="https://bittrex.com/">Bittrex</a> by following the instructions <a href="https://support.bittrex.com/hc/en-us/articles/115000199631-How-do-I-deposit-coins-to-Bittrex-">on this page</a>. To send Bitcoin on Coinbase, go to the Accounts page and add the Bittrex Bitcoin address and amount.</li><li>Buy Neo on Bittrex by going to the Bitcoin Markets page and searching for Neo.</li><li>Wait for the price to go up.</li><li>Sell Neo for Bitcoin in reverse of step 3.</li><li>Send Bitcoin over to Coinbase.</li><li>Sell Bitcoin on Coinbase similarly to selling Ethereum as previously described.</li></ol><h4>Bitcoin — the Hard Way</h4><ol><li>Purchase $600 worth of Bitcoin on Coinbase (see above).</li><li>Send Bitcoin to <a href="https://gdax.com/">GDAX</a> (a Coinbase exchange) by following the instructions <a href="https://support.gdax.com/customer/en/portal/articles/2430129-transferring-funds-between-gdax-and-coinbase">on this page</a>.</li><li>Prepare the Bitcoin for sale by setting up a Limit Sell in the amount that I’m happy with (my go-to is +18%).</li><li>Wait for the price to go up.</li><li>Withdraw USD from GDAX directly into my bank account.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aeb35c47fe6f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/actually-making-money-by-trading-cryptocoins-aeb35c47fe6f">Actually Making Money by Trading Cryptocoins</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Creating a Daily Comic Slack Bot]]></title>
            <link>https://medium.com/the-academy/creating-a-daily-comic-slack-bot-f740b9e520a2?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/f740b9e520a2</guid>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[dilbert]]></category>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[slack]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Sun, 18 Dec 2016 03:45:21 GMT</pubDate>
            <atom:updated>2016-12-18T03:45:21.084Z</atom:updated>
            <content:encoded><![CDATA[<p>When it comes to task automation, few tools are as powerful as bots. Yet, not all bots have to solve serious business problems. Some can be fun. In this article, I am going to show you how to deploy a simple bot that posts a daily Dilbert comic to a Slack channel. <a href="https://github.com/ilyakam/dilbertbot/">The source code is available on GitHub.</a></p><h4>Slack Integration</h4><ol><li>Create a new <em>Incoming WebHook</em> on the <a href="https://slack.com/apps/manage/custom-integrations">Custom Integrations page</a></li><li>Copy and save the Webhook URL<br>It should start with <a href="https://hooks.slack.com/">https://hooks.slack.com</a><br>You’ll need it for the commands below</li></ol><h4>Deploy Bot with Zeit’s Now</h4><ol><li>Download “Δ now” from <a href="https://zeit.co/">Zeit.co</a><br>Alternatively, install it with npm (npm install now -g)</li><li>Deploy the application with the environment flag<br>now -e SLACK_WEBHOOK_URL=&lt;URL from #2 above&gt;</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/662/1*0nEN-wyOi4Uj15QvuWVFsA.png" /><figcaption>And that’s it!</figcaption></figure><p>If you want to deploy my version of the bot, so that it posts to the <em>#dilbert</em> channel every day at 15:00 UTC (7am PST / 10am EST), you can do it with one simple command:</p><pre>now deploy \<br>  -e SLACK_WEBHOOK_URL=<strong>[CHANGE ME]</strong> \<br>  <a href="https://github.com/ilyakam/dilbertbot">https://github.com/ilyakam/dilbertbot</a></pre><p>Remember to substitute [CHANGE ME] with your SLACK_WEBHOOK_URL in the command above.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f740b9e520a2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/creating-a-daily-comic-slack-bot-f740b9e520a2">Creating a Daily Comic Slack Bot</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Exploring Jasmine’s Custom Asymmetric Equality Tester]]></title>
            <link>https://medium.com/the-academy/exploring-jasmines-custom-asymmetric-equality-tester-18e07f08a1e1?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/18e07f08a1e1</guid>
            <category><![CDATA[jasmine]]></category>
            <category><![CDATA[angularjs]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[front-end-development]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Sat, 26 Nov 2016 10:40:35 GMT</pubDate>
            <atom:updated>2016-11-26T10:40:34.889Z</atom:updated>
            <content:encoded><![CDATA[<p>There’s a <a href="https://jasmine.github.io/2.5/introduction.html#section-Custom_asymmetric_equality_tester">hidden gem in the Jasmine testing framework</a> called asymmetricMatch. Since I could not find it being mentioned anywhere on the blogosphere, I have decided that it deserves a dedicated post of its own.</p><p>The asymmetric equality tester allows us to test any custom expectation at a later time. To illustrate this point, I have combined three test scenarios in <a href="https://github.com/ilyakam/jasmine-asymmetric-tester">a single project on GitHub</a>. Feel free to clone the entire repository and to follow along.</p><h4>Angular-Foundation Modal Resolver</h4><p>When opening modals with Angular-Foundation, <a href="https://pineconellc.github.io/angular-foundation/#/modal">the docs specify</a> that each resolved item needs to be defined using a function that returns a value. In order to avoid holes in the test coverage report, I used to resort to mocking the $modalProvider in its entirety. However, that would result in having huge spec files that would needlessly test a big chunk of the Angular-Foundation module.</p><p>With Jasmine’s asymmeticMatch function, it is easy to only test the code that matters most. Here’s how:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e754408dcb12dd32ffe39cd7fa006474/href">https://medium.com/media/e754408dcb12dd32ffe39cd7fa006474/href</a></iframe><h4>Angular UI-Router State Resolver</h4><p>Similarly to the modal resolver above, the <a href="https://ui-router.github.io/ng1/">Angular UI-Router</a> also resolves values asynchronously via a resolve function. The added complexity here is that it all happens within the configuration block, meaning that we also have to inject a mock provider. In fact, <a href="http://nikas.praninskas.com/angular/2014/09/27/unit-testing-ui-router-configuration/#resolve">Nikas Praninskas wrote an excellent tutorial</a> on how to unit test the UI-Router resolve method. By using the asymmetric equality tester, I argue that it’s possible to achieve the same result with a lot less boilerplate code. Take a look at the following Gist:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/0bf01eb8e957088f8afe130637a8cf3b/href">https://medium.com/media/0bf01eb8e957088f8afe130637a8cf3b/href</a></iframe><h4>ngRoute Route Resolver</h4><p>Since the <a href="https://docs.angularjs.org/api/ngRoute/service/$route">ngRoute</a> module closely resembles the previous example, I have included it below as well. Here’s what testing its resolve function would look like when combined with asymmetricMatch:</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/8ad2f2307213fa46c90a595174efb149/href">https://medium.com/media/8ad2f2307213fa46c90a595174efb149/href</a></iframe><p>Have you used Jasmine’s asymmetricMatch in other scenarios? Please share your experience in the comments section below. Pull requests are welcome.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=18e07f08a1e1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/exploring-jasmines-custom-asymmetric-equality-tester-18e07f08a1e1">Exploring Jasmine’s Custom Asymmetric Equality Tester</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[OS X Modifier Keys on a PC Keyboard]]></title>
            <link>https://medium.com/the-academy/os-x-modifier-keys-on-a-pc-keyboard-d31d28a9ac9?source=rss----7b9861ba4a01---4</link>
            <guid isPermaLink="false">https://medium.com/p/d31d28a9ac9</guid>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[keyboard]]></category>
            <category><![CDATA[mac]]></category>
            <category><![CDATA[shortcuts]]></category>
            <category><![CDATA[os-x]]></category>
            <dc:creator><![CDATA[Ilya Kaminsky]]></dc:creator>
            <pubDate>Wed, 03 Aug 2016 17:26:06 GMT</pubDate>
            <atom:updated>2016-08-03T17:26:06.447Z</atom:updated>
            <content:encoded><![CDATA[<p>Ever tried using the copy/paste shortcuts on your Mac with a PC keyboard? If so, you must have realized that Control+C does not work, and the Command key is nowhere to be found. Some people are OK with the default mapping of the Windows key to replace the Command key. In case you are not, here’s how to remap the modifier keys so that their layout is all too familiar to you:</p><ol><li>Go to Preferences » Keyboard » Keyboard Tab » Modifier Keys…</li><li>Select the PC keyboard from the dropdown above</li><li>Map Option (⌥) to ⌘ Command</li><li>Map Command (⌘) to ⌥ Option</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SRhi_Cx3ZqhuFPuI93dslQ.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d31d28a9ac9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-academy/os-x-modifier-keys-on-a-pc-keyboard-d31d28a9ac9">OS X Modifier Keys on a PC Keyboard</a> was originally published in <a href="https://medium.com/the-academy">The Academy</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>