<?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 Ankit Sachan on Medium]]></title>
        <description><![CDATA[Stories by Ankit Sachan on Medium]]></description>
        <link>https://medium.com/@ankitsachan?source=rss-d0597548905e------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*w3SWDcMZW4KyPoFq.</url>
            <title>Stories by Ankit Sachan on Medium</title>
            <link>https://medium.com/@ankitsachan?source=rss-d0597548905e------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 21:29:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ankitsachan/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[School Taught You to Be a Library. AI Just Made Libraries Free.]]></title>
            <link>https://medium.com/@ankitsachan/school-taught-you-to-be-a-library-ai-just-made-libraries-free-3352f787790c?source=rss-d0597548905e------2</link>
            <guid isPermaLink="false">https://medium.com/p/3352f787790c</guid>
            <category><![CDATA[future-of-work]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[self-improvement]]></category>
            <dc:creator><![CDATA[Ankit Sachan]]></dc:creator>
            <pubDate>Tue, 24 Feb 2026 14:38:40 GMT</pubDate>
            <atom:updated>2026-02-24T14:38:40.515Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8z_xRD1vboOCRmPT5H2zyA.png" /></figure><h4>What happens when the thing you spent 16 years learning to do becomes a commodity overnight?</h4><p>Here’s a question that should keep every educator up at night:</p><p><strong>Why do we spend 16+ years stuffing information into human heads when a machine can now access all of it in under a second?</strong></p><p>Not “some” of it. Not “a simplified version.” Virtually all recorded human knowledge — synthesized, explained, applied to your specific problem — available instantly to anyone with a phone.</p><p>That’s not a small upgrade. That changes the entire game.</p><p>And I think it exposes something uncomfortable: <strong>the system we call “education” was never really about making you smart. It was about making you a reliable storage device.</strong></p><p>Let me explain.</p><h3>The 4,000-Year-Old Copy-Paste Machine</h3><p>Go back to the very first schools. Literally the first ones.</p><p>Around 2000 BCE in ancient Mesopotamia, they had “tablet houses” — the world’s earliest classrooms. What did students do there? They copied. All day. They copied word lists. They copied math tables. They copied proverbs. A student text from that era reads:</p><blockquote><em>“I recited my tablet, ate my lunch, prepared my new tablet, wrote it, finished it.”</em></blockquote><p>That was school. Copy, memorize, repeat.</p><p>Now fast-forward 4,000 years. Walk into a modern classroom. A teacher stands at the front. Students sit in rows. Information flows one direction. There’s a test on Friday to check if the copy-paste worked.</p><p>The technology changed (clay tablets to iPads). The architecture didn’t.</p><p><strong>We’re still running Mesopotamian software on 21st-century hardware.</strong></p><figure><img alt="Split illustration comparing an ancient Mesopotamian classroom with a modern classroom, showing identical layouts across 4000 years" src="https://cdn-images-1.medium.com/max/1024/1*TGD0N-Y2gk0IRj8CisuqUQ.png" /></figure><h3>Every “Revolution” Hit the Same Wall</h3><p>Here’s what’s wild. This isn’t the first time a new technology threatened to make the old way of learning obsolete. It’s happened at least three times before.</p><p><strong>And every single time, people freaked out about the same thing.</strong></p><figure><img alt="Timeline illustration showing four knowledge disruptions: clay tablets, Greek scrolls, printing press, and computers" src="https://cdn-images-1.medium.com/max/1024/1*QVx4h_1N-TwJr8vaMFkkBg.png" /></figure><h4>Round 1: Writing (400 BCE)</h4><p>As written texts proliferated across ancient Greece, Socrates — yes, *<em>that</em>* Socrates — had a meltdown. He told a story about an Egyptian god who presented writing to a king, claiming it would make people wiser. The king shot back:</p><blockquote>“You’ve discovered an elixir not of memory, but of reminding. You give your students the appearance of wisdom, not its reality.”</blockquote><p>Translation: “They’ll think they’re smart because they read stuff, but they won’t actually <em>understand</em> anything.”</p><p>Sound familiar? Swap “writing” for “ChatGPT” and that quote could’ve been posted online last week.</p><h4>Round 2: The Printing Press (1440s)</h4><p>Before Gutenberg, Europe had maybe 30,000 books total. Fifty years after the printing press? <strong>20 million volumes.</strong></p><p>People panicked. An abbot named Trithemius wrote an actual book titled <strong><em>In Praise of Scribes</em></strong> arguing that printed books would destroy real learning. (The book was, of course, printed. He saw no irony.)</p><p>The printing press was supposed to kill universities. Instead, it killed the monopoly on knowledge held by a small elite. Universities adapted. Eventually.</p><h4>Round 3: The Internet (2000s)</h4><p>In 2012, Udacity co-founder Sebastian Thrun predicted only 10 universities would survive within 50 years. Clayton Christensen said half of American universities would be bankrupt by 2027.</p><p>What actually happened? MOOC completion rates were <strong>3–6%</strong>. Wikipedia became the greatest encyclopedia ever made — but it turns out having access to information and actually <em>learning</em> from it are two very different things.</p><h4>The Pattern</h4><p>Here’s what I want you to see:</p><p>Every time a new technology makes knowledge more accessible, two things happen:</p><p>1. <strong>The access problem gets solved spectacularly.</strong> Books become cheap. Information goes online. Knowledge becomes free.</p><p>2. <strong>The comprehension problem stays exactly the same.</strong> People still need time, effort, guidance, and struggle to actually <strong><em>understand</em></strong> things.</p><p>We keep confusing “access to information” with “acquisition of understanding.” They’re not the same thing. They’ve never been the same thing.</p><p><strong>But here’s where AI breaks the pattern.</strong></p><h3>Why AI Is Different (For Real This Time)</h3><p>Every previous technology — writing, printing, Google — made information <strong><em>available</em></strong>. You still had to find it, read it, synthesize it, and make sense of it yourself.</p><p>AI doesn’t just hand you information. It synthesizes, analyzes, explains, and applies it <strong><em>for you</em></strong>.</p><p>Ask Google: “What are the key arguments in Kahneman’s <strong><em>Thinking, Fast and Slow</em></strong>?”</p><p>You get 10 links. You click, read, and figure it out.</p><p>Ask an AI the same question?</p><p>You get a clear, structured summary tailored to your level, with follow-up questions if you want to go deeper. In 10 seconds.</p><p>That’s not the same upgrade as “books got cheaper.” That’s a fundamentally different relationship with knowledge.</p><p>For the first time in history, the bottleneck isn’t “<strong>Can I access the knowledge?”</strong></p><p>It’s “<strong>Can I figure out the right question to ask?”</strong>**</p><h3>The Factory That Forgot to Update</h3><p>Here’s what makes this awkward.</p><p>In 1916, a prominent education administrator named Ellwood Cubberley described schools this way:</p><p><em>“Our schools are, in a sense, factories, in which the raw products (children) are to be shaped and fashioned into products to meet the various demands of life.”</em></p><p>He wasn’t being critical. He was <strong><em>proud</em></strong> of this.</p><p>The Prussian education model — the one that spread globally and still dominates today — was explicitly designed to produce:</p><p>- People who follow instructions</p><p>- People who show up on time</p><p>- People who can process standardized information</p><p>- People who don’t question authority too much</p><p>This was great… for 1850. When factories needed literate, obedient workers. When knowledge was scarce and the economy rewarded people who could store and reproduce it reliably.</p><p>But the economy isn’t asking for that anymore. And AI just made the storage-and-reproduction part trivially easy.</p><p>So what <strong><em>is</em></strong> it asking for?</p><h4>The Big Flip</h4><p>A Harvard/BCG study in 2023 tested something interesting. They gave 758 management consultants access to GPT-4 and tracked what happened.</p><p>The results were wild:</p><p>- On tasks the AI was good at, consultants with AI finished <strong>25% faster</strong> with <strong>40% higher quality.</strong></p><p>- On tasks <strong><em>outside</em></strong> AI’s strengths, consultants who trusted the AI blindly performed <strong>19% worse</strong> than those without AI at all.</p><p>The difference between the winners and losers wasn’t who was smarter. It wasn’t who knew more.</p><p><strong>It was who had better judgment about when to trust the AI and when to override it.</strong></p><p>That’s the flip.</p><p>The old world rewarded: <strong><em>How much do you know?</em></strong></p><p>The new world rewards: <strong><em>How well can you think when machines know everything?</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Gjoys45Kisi5P_5c83UiPg.png" /></figure><h3>What Actually Matters Now (The Five Things)</h3><p>If knowledge is no longer the scarce resource, what is? Based on everything we’re seeing across medicine, law, tech, science, and education — it comes down to five things:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3pAEvvZhYtxdUr03-XWLwQ.png" /></figure><h4>1. Asking the right questions</h4><p>AI is spectacular at answering questions. It’s terrible at knowing which questions matter.</p><p>Einstein supposedly said: “<strong><em>If I had an hour to solve a problem, I’d spend 55 minutes defining the problem and 5 minutes solving it.</em></strong>”</p><p>That ratio just got way more extreme. AI handles the 5 minutes. The 55 minutes of figuring out what problem you’re actually solving? That’s all you.</p><p>The rise of “<strong><em>prompt engineering</em></strong>” is literally the professionalization of question-asking. But it goes deeper than prompts — it’s about framing the right problems, seeing what others miss, and knowing what to point the machine at.</p><h4>2. Judgment (not knowledge)</h4><p>A doctor who memorized every disease isn’t as valuable as a doctor who can look at a patient, feel that something’s off, and say “I know the textbook says X, but something else is going on here.”</p><p>That gut-level, experience-built judgment — the ability to notice what’s missing, what doesn’t fit, what the data isn’t telling you — that’s the new premium skill.</p><p>And it’s the one thing you can’t get from AI. Because it’s built through years of <strong><em>being wrong</em></strong>, adjusting, and building better instincts.</p><h4>3. Connecting dots across fields</h4><p>AI is the ultimate specialist. It knows everything about narrow domains. What it’s not great at (yet) is seeing unexpected connections between distant fields.</p><p>The best innovations almost always come from someone who saw a pattern in one domain and applied it somewhere completely different. Darwin borrowed from economics. Steve Jobs borrowed from calligraphy. The mRNA vaccine was built by people who connected immunology, RNA biology, and lipid nanoparticle chemistry — fields that didn’t usually talk to each other.</p><p>In a world of AI specialists, the generalist who connects dots becomes incredibly valuable.</p><h4>4. Knowing what you don’t know</h4><p>This sounds philosophical, but it’s extremely practical.</p><p>AI confidently generates wrong answers all the time. It hallucinates citations. It makes up statistics. It states uncertain things with perfect confidence.</p><p>The skill of the future isn’t “being right.” It’s knowing <em>when you might be wrong </em>— and knowing when the AI might be wrong too.</p><p>I’d call this “epistemic awareness,” but that sounds too fancy. Let’s just call it <strong>intellectual honesty on steroids.</strong> Knowing the limits of your own knowledge AND the limits of the machine’s knowledge. That’s rare. That’s valuable.</p><h4>5. Working <em>with</em> the machine, not against it or under it</h4><p>The BCG study found two winning strategies:</p><p>- <strong>Centaurs:</strong> People who clearly divided work — “AI does this part, I do that part”</p><p>- <strong>Cyborgs:</strong> People who deeply wove AI into every step, constantly going back and forth</p><p>Both crushed it. What failed? People who just handed everything to AI and hit “accept.”</p><p>The new skill isn’t using AI or avoiding AI. It’s <strong><em>collaborating</em></strong> with it — like a musician working with a really talented but slightly unreliable bandmate.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9icMu7DDICEr7LtfEMFmPg.png" /></figure><h4>The Uncomfortable Middle</h4><p>Now, I want to be honest about something. This is not a “schools are dead, learn on YouTube” article.</p><p>There are strong reasons traditional education sticks around:</p><p><strong>You can’t judge what you don’t understand.</strong> All those “higher-order skills” I described? They actually require a foundation of knowledge. You can’t ask good questions about biology if you don’t know any biology. You can’t have good judgment about code if you’ve never written code. The base matters.</p><p><strong>Struggle is where learning lives.</strong> Research shows that making things <strong><em>harder</em></strong> (what psychologists call “desirable difficulty”) actually improves learning. If AI removes all the struggle, it might remove all the learning too.</p><p><strong>Schools do more than teach.</strong> They socialize kids. They provide structure. They build identity. They create shared cultural experiences. AI doesn’t replace any of that.</p><p><strong>Every tech revolution has been over-hyped in the short term.</strong> Remember when MOOCs were going to replace universities? When TV was going to revolutionize education? When calculators were going to free up math class for “higher-order thinking?” The revolution is always slower and messier than the hype suggests.</p><p>So the honest take isn’t “education is dead.” It’s something more nuanced:</p><p><strong>Education was built to solve a knowledge-scarcity problem. AI just solved that problem. Now education needs to solve a different one: teaching people to think well in a world drowning in knowledge.</strong></p><h4>What Schools Could Look Like</h4><p>If we take this seriously, education flips in some pretty radical ways:</p><p><strong>Problem-first, not prerequisite-first.</strong> Instead of “learn math for 4 years, then maybe apply it” — start with a real problem, and learn math because you need it to solve the thing in front of you. AI can fill in gaps on the fly.</p><p><strong>Test judgment, not recall.</strong> Instead of “what year did the French Revolution start?” — try “here’s an AI-generated analysis of the French Revolution. Find the three things it got wrong and explain why they matter.” That tests something AI can’t do for you.</p><p><strong>Teach metacognition explicitly.</strong> “How do you know what you know?” should be a core class, not a philosophical curiosity. Confidence calibration, bias recognition, source evaluation — these are survival skills now.</p><p><strong>Make learning continuous, not front-loaded.</strong> The idea that you learn for 16–22 years and then “work” for 40 is absurd in a world where 44% of core skills are expected to be disrupted every five years (World Economic Forum, 2023). Learning needs to be woven into work, not separated from it.</p><p><strong>Use AI as the teacher’s superpower, not its replacement.</strong> The best evidence (Khan Academy, multiple university studies) shows AI works best when paired with human instruction — not as a replacement for it. The teacher becomes the coach, the guide, the person who asks “but do you actually understand what the AI just told you?”</p><h4>The Real Shift</h4><p>Let me leave you with the simplest version of all this:</p><p>For 4,000 years, education’s job was to turn people into good <strong>storage devices</strong> for knowledge. That made sense. Knowledge was expensive. Having it in your head was a genuine advantage.</p><p>AI just made storage free.</p><p>The new job of education — the one we’re still figuring out, the one no one has fully cracked — is to turn people into good <strong>thinking devices.</strong> People who can judge, question, connect, and create in a world where raw knowledge is as cheap as tap water.</p><p>That’s not a small change. It’s the biggest shift in learning since… well, since Socrates complained about writing ruining everything.</p><p>And if history is any guide? He was about 30% right.</p><p>The question is whether we’ll be smart enough — actually <strong><em>smart</em></strong>, not just informed — to handle the other 70%.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yBF7hp6ank4BdoAd5Bhxbg.png" /></figure><p><em>If you found this interesting, I’d love to hear what you think. Are schools preparing people for the AI era? Or are we still running 4,000-year-old software? Drop a comment or reach out.</em></p><p><strong>Further Reading:</strong></p><p>- Ethan Mollick, <em>Co-Intelligence </em>(2024) — Best practical guide to human-AI collaboration</p><p>- David Epstein, <em>Range </em>(2019) — Why generalists beat specialists</p><p>- Sal Khan, <em>Brave New Words</em> (2024) — How AI could fix education</p><p>- Dell’Acqua et al., “Navigating the Jagged Technological Frontier” (2023) — The Harvard/BCG study referenced above</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3352f787790c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Android Studio + Kotlin + Open CV]]></title>
            <link>https://medium.com/@ankitsachan/android-studio-kotlin-open-cv-16d75f8d9969?source=rss-d0597548905e------2</link>
            <guid isPermaLink="false">https://medium.com/p/16d75f8d9969</guid>
            <dc:creator><![CDATA[Ankit Sachan]]></dc:creator>
            <pubDate>Mon, 14 Aug 2023 12:35:38 GMT</pubDate>
            <atom:updated>2023-08-14T12:35:38.833Z</atom:updated>
            <content:encoded><![CDATA[<p>so this is if you are getting frustrated with how to integrate OpenCV 4.x.x with Android Studio.</p><ol><li><strong>Download Open CV SDK (</strong><a href="https://opencv.org/releases/"><strong>https://opencv.org/releases/</strong></a><strong>)</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/777/1*3HC8pfhvxi4u1q6S68zgbw.png" /></figure><p><strong>2.</strong> <strong>Unzip the SDK</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/496/1*5NdYktcdSouzJMPsMlHkyA.png" /></figure><p><strong>3. Create an Android project</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/899/1*33A2mAZYXAxTu-PCLNWNUg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/900/1*fDy10iOnxLoODtarXuPP3w.png" /></figure><p><strong>4. Add OpenCV SDK as a dependency</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/717/1*lyVavu8_iNXYdZ5txoBg0w.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/656/1*w64UGI-6RdHw96a3oNxqvw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/903/1*CRGlelz_gcXBH_26j-vgXg.png" /></figure><p>As soon as you’ll add you’ll see this error</p><pre>Namespace not specified. Please specify a namespace in the module&#39;s build.gradle file like so:<br><br>android {<br>    namespace &#39;com.example.namespace&#39;<br>}</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e6UXwys09nk_P3QL36FR0g.png" /></figure><p><strong>5. Let&#39;s solve this error</strong></p><p>open android project as Project:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/236/1*Mn_F3S464fLNubk0BshL7Q.png" /></figure><p>access the build.gradle file of opencv</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/290/1*dZcD-a4Quq0zfRAp39elpw.png" /></figure><p>add namespace ‘org.opencv’ under android</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UNDK7jD-oG2M8RcYaTfTDg.png" /></figure><p>sync the gradle by tapping on Try again</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hOVRWaecWvgXWlgGJKZsUg.png" /></figure><p>you should see something like this</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Jom10_ewxJPVUG8-BN5Z5Q.png" /></figure><p><strong>6. Let&#39;s add opencv as a dependency to the main app</strong></p><p>open project structure</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/439/1*bOHG0fkOCr72N1ox9FmWRA.png" /></figure><p>add ‘opencv’ as a dependency to the main app</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/950/1*Sd1If5eqP9HEclsUJJgO7g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/597/1*nhCmDXWKcrhFFtAR5kKPZw.png" /></figure><p><strong>7. Run the app in debug mode you should see this error:</strong></p><pre>error: package org.opencv.engine does not exist<br>import org.opencv.engine.OpenCVEngineInterface;<br>                        ^</pre><p>to fix this we need to modify the build feature to support aidl</p><p><a href="https://github.com/opencv/opencv/pull/23447#issuecomment-1523533632">AGP 8.0 build.gradle namespace and aidl buildFeature requirement added by onuralpszr · Pull Request #23447 · opencv/opencv</a></p><p>open build.gradle of ‘opencv’ and add</p><pre>buildFeatures {<br>        aidl true<br>}</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rbZMlDJOdJOTt6zcFL4N5w.png" /></figure><p><strong>8. After this you should get this error</strong></p><pre>import org.opencv.BuildConfig;</pre><p>Fix:</p><p><a href="https://forum.opencv.org/t/an-exercise-in-frustration/12964/4?u=ankit_sachan">An exercise in frustration</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OaneMdgXM2SqdrU7oKVEOQ.png" /></figure><p><strong>9. Let&#39;s check if OpenCV is successfully added to the project or not</strong></p><p>Add these lines to MainActivity.kt</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BYe_GFXNRZiq6sUWUiTZgA.png" /></figure><p>You should see this output.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Z-36pDXr8acK6o_8kG0iEA.png" /></figure><p>Hope this helps you.</p><p>Happy Coding.</p><p><a href="https://medium.com/u/d19d774b0c8">Satya Mallick</a>: Please get the documentation updated</p><p><strong>Acknowledgements</strong>:</p><p><a href="https://forum.opencv.org/t/an-exercise-in-frustration/12964">An exercise in frustration</a></p><p><a href="https://github.com/opencv/opencv/pull/23447">https://github.com/opencv/opencv/pull/23447</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=16d75f8d9969" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[iOS Application Localization]]></title>
            <link>https://medium.com/@ankitsachan/ios-application-localization-47ce83c0872b?source=rss-d0597548905e------2</link>
            <guid isPermaLink="false">https://medium.com/p/47ce83c0872b</guid>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[weblate]]></category>
            <category><![CDATA[ios]]></category>
            <category><![CDATA[github]]></category>
            <category><![CDATA[localization]]></category>
            <dc:creator><![CDATA[Ankit Sachan]]></dc:creator>
            <pubDate>Fri, 31 Jan 2020 05:27:13 GMT</pubDate>
            <atom:updated>2020-01-31T05:27:13.522Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/489/1*w4v-KFiSmx8RZHBlRgiNiA.png" /></figure><h3>What is localization?</h3><p>Localization is the process of translating your app into multiple languages. Tailoring your app’s interface to meet varying cultural and language expectations is a must. Users want to feel that the apps they download are created natively for them, and having obviously incorrect terminology or culturally insensitive content is a huge turn-off.</p><h3>What we know?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/591/1*_dy7IuQG59rRAVdBbPaqAQ.png" /></figure><p>So the above diagram shows the typical workflow which a developer follows while localizing an iOS application. This process can be further complicated depending upon the organization of string files. So we can clearly see lots of repetitive work happening and thus there is an opportunity to optimize via automation.</p><h3>A better proposition</h3><p>So we can have two ways in which we can improve this process:</p><h4>LocalizationKit:</h4><p>This is an open-source library and offers some great benefit. <a href="https://github.com/willpowell8/LocalizationKit_iOS">https://github.com/willpowell8/LocalizationKit_iOS</a></p><h4>Pros</h4><ul><li>Update the strings from the web dashboard and strings will be updated in app.</li><li>Strings will be automatically pushed to the web dashboard</li><li>free</li></ul><h4>Cons</h4><ul><li>Not suitable for enterprise-level products</li></ul><h4>Weblate:</h4><p>This is an open-source product as well. <a href="https://weblate.org/en/">https://weblate.org/en/</a></p><h4>Pros</h4><ul><li>Automatically pulls the strings from the configured repository</li><li>Provides machine translation</li><li>Automatically pushes the code or raise a PR to the configured repository</li><li>Supports all major version control system</li><li>Can be used in enterprise products</li><li>Can be deployed on internal servers</li><li>Translators can be invited to a web dashboard</li><li>Continuous Integration support</li><li>Notifications via email on change of string files</li></ul><h4>Cons</h4><ul><li>No OTA support</li></ul><h3>Conclusion</h3><p>Adopting these tools we ensure that we automate the repetitive tasks we perform during application localization. Please reach out on ankit.sachan16@gmail.com if you need help on above-mentioed tools.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=47ce83c0872b" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>