<?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 Kagadevishal on Medium]]></title>
        <description><![CDATA[Stories by Kagadevishal on Medium]]></description>
        <link>https://medium.com/@kagadevishal?source=rss-b56df248cb69------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*NnxMgdARToVOQoCYxRChjg.jpeg</url>
            <title>Stories by Kagadevishal on Medium</title>
            <link>https://medium.com/@kagadevishal?source=rss-b56df248cb69------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 22:56:42 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@kagadevishal/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[The Scammer Who Quizzed Me on Mathematics]]></title>
            <link>https://medium.com/illumination/the-scammer-who-quizzed-me-on-mathematics-9c901275754e?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/9c901275754e</guid>
            <category><![CDATA[humans]]></category>
            <category><![CDATA[humanity]]></category>
            <category><![CDATA[scam]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Wed, 01 Jul 2026 15:55:57 GMT</pubDate>
            <atom:updated>2026-07-02T08:47:07.183Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/596/1*9R5PHkzDtATi82ZYXEAq7A.png" /><figcaption><a href="https://www.makios.com/">https://www.makios.com/</a></figcaption></figure><h3>How a scam call turned into an unexpected conversation about mathematics, PhDs, and the harsh realities of life.</h3><p>We all know the drill. Your phone rings, an unknown number flashes on the screen, and the moment you answer, an automated voice or an urgent-sounding operator tells you that your money is gone.</p><p>Recently, I received one of these calls. It started exactly like you would expect, but it ended in a way I will never forget.</p><p><strong>The Setup</strong></p><p>I was in the middle of working when my phone buzzed. On the other end, a voice informed me that €500 had just been deducted from my PayPal account. <em>Did I authorize this? Did I want to speak to customer support?</em></p><p>I pressed the button to speak to support, curious to see where this was going.</p><p>The man who took over the line immediately launched into the classic scammer script. He told me that my money had been lost in some crypto transaction because a hacker had compromised my account. He asked if I had recently connected to a public Wi-Fi network. When I said yes, he seized on it: <em>“That is how they got you.”</em></p><p>The “solution,” of course, was to go to the Google Play Store and download a specific app. I knew right away what the app was: a remote desktop application designed to give them full access to my phone.</p><p>I also noticed something else. Like me, the man on the other end of the line was from India. I recognized his accent instantly, and the whole charade suddenly felt very transparent.</p><h3>Breaking a Script</h3><p>Instead of playing along, panicking, or just hanging up, I decided to be direct, but respectful.</p><blockquote><em>“Look,” I told him. “I am doing my PhD in Artificial Intelligence. I know exactly what you are doing, and you know exactly what you are doing. I am being respectful to you. Let’s not waste your time, and let’s not waste mine.”</em></blockquote><p>I caught him completely off guard. Scammers are trained to deal with fear, compliance, anger, and people slamming the phone down. They are not trained for calm, respectful honesty.</p><p>For a second, the line was quiet. When he spoke again, the scammer persona had vanished.</p><p><em>“You’re doing your PhD?”</em> he asked, his tone suddenly defensive but curious. <em>“I also did a PhD. In mathematics.”</em></p><h3>The Unexpected Math Quiz</h3><p>What happened next was surreal. To prove his credentials, or perhaps just to reclaim some intellectual pride, this anonymous scammer decided to test me.</p><p><em>“Okay, you’re doing a PhD,”</em> he challenged. <em>“Let me ask you some mathematics questions.”</em></p><p>For the next few minutes, sitting at my desk, I was grilled on matrix mathematics and vector calculus by a man who had just tried to steal €500 from me. And the questions weren’t surface-level trivia. They were deep, foundational, and incredibly well-thought-out.</p><p>I answered the ones I knew. I stumbled on the ones I didn’t. When I didn’t know the answer, <em>he explained the solutions to me</em>. He walked me through the formulas with the patience of an academic. It became completely obvious that he was telling the truth: this man was highly educated.</p><h3>The Confession</h3><p>After our impromptu math seminar wrapped up, the atmosphere between us had completely shifted. We were no longer predator and prey; we were just two Indian academics talking over the phone.</p><p>I decided to push my luck and ask him the one question on my mind.</p><p>“I spoke to you honestly, and I treated you with respect,” I said. “Can you please tell me honestly… were you actually going to do the thing you called me for? Did you have another intention?”</p><p>I told him he didn’t have to explain himself, just a simple yes or no.</p><p><em>“Yes,”</em> he admitted.</p><p>Then, he elaborated. He told me that sometimes, people find themselves in incredibly difficult situations. Sometimes, you have to do things you don’t want to do, even if it means harming others.</p><p>He told me he used to be a mathematics professor. <em>“But some shit happened,”</em> he said quietly. <em>“And I ended up here.”</em></p><h3>A Human Connection</h3><p>It is easy to look at cybercriminals as faceless villains. And make no mistake: scamming people is wrong, and it destroys livelihoods. But hearing the quiet resignation in his voice was a sobering reminder of the complex, often tragic human realities operating behind these massive scam call centers. Here was a man with a brilliant, educated mind, capable of teaching high-level calculus, reduced to reading a script to steal money from strangers.</p><p>Before we hung up, I didn’t lecture him. I just wished him well.</p><p>“I understand,” I told him. “I hope you find something good to do with your talent, and I hope you find a way out of this.”</p><p>His final words to me weren’t a threat or a slick sales pitch. They were the words of a former professor offering advice to a student.</p><p><em>“Study very hard,”</em> he told me. <em>“Focus on your studies. Have a good day.”</em></p><p>I kept my money that day, but I walked away with something far more valuable: a reminder that humanity, empathy, and even vector calculus can show up in the most unexpected places.</p><p>Have you ever had an unexpected human interaction in a bizarre situation? Share your thoughts in the comments.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9c901275754e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/illumination/the-scammer-who-quizzed-me-on-mathematics-9c901275754e">The Scammer Who Quizzed Me on Mathematics</a> was originally published in <a href="https://medium.com/illumination">ILLUMINATION</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Quotes That Found Me When I Needed Them Most]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/activated-thinker/quotes-that-found-me-when-i-needed-them-most-87920c9dc631?source=rss-b56df248cb69------2"><img src="https://cdn-images-1.medium.com/max/2600/0*Ew6Nyvqnvy6Z4mST" width="6000"></a></p><p class="medium-feed-snippet">Moments when words understood me better than people</p><p class="medium-feed-link"><a href="https://medium.com/activated-thinker/quotes-that-found-me-when-i-needed-them-most-87920c9dc631?source=rss-b56df248cb69------2">Continue reading on Activated Thinker »</a></p></div>]]></description>
            <link>https://medium.com/activated-thinker/quotes-that-found-me-when-i-needed-them-most-87920c9dc631?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/87920c9dc631</guid>
            <category><![CDATA[motivation]]></category>
            <category><![CDATA[life]]></category>
            <category><![CDATA[quotes]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Tue, 16 Dec 2025 11:45:14 GMT</pubDate>
            <atom:updated>2026-01-02T07:08:37.433Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Temperature is all you need — How Temperature controls creativity.]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://ai.plainenglish.io/temperature-is-all-you-need-how-temperature-controls-creativity-33e1ed270ca5?source=rss-b56df248cb69------2"><img src="https://cdn-images-1.medium.com/max/730/1*kfiphh4gUx1cx32wavnmdw.png" width="730"></a></p><p class="medium-feed-snippet">A few days ago, one of my friend from management told me that he is using LLM&#x2019;s in Google Studio (GEMINI)&#xA0;, and also ChatGPT sometimes. He&#x2026;</p><p class="medium-feed-link"><a href="https://ai.plainenglish.io/temperature-is-all-you-need-how-temperature-controls-creativity-33e1ed270ca5?source=rss-b56df248cb69------2">Continue reading on Artificial Intelligence in Plain English »</a></p></div>]]></description>
            <link>https://ai.plainenglish.io/temperature-is-all-you-need-how-temperature-controls-creativity-33e1ed270ca5?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/33e1ed270ca5</guid>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[deterministic]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[creativity]]></category>
            <category><![CDATA[temperature]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Sun, 14 Dec 2025 07:08:22 GMT</pubDate>
            <atom:updated>2026-01-31T07:20:29.672Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Why Schools Must Rethink Learning in the Age of AI?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@kagadevishal/why-schools-must-rethink-learning-in-the-age-of-ai-a4b478203888?source=rss-b56df248cb69------2"><img src="https://cdn-images-1.medium.com/max/1940/1*y-8BtLTk8rkAZe_BjNbAKw.jpeg" width="1940"></a></p><p class="medium-feed-snippet">A few days ago, Google released its new image generation model called Nano Banana Pro. This model is extremely powerful and can do amazing&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@kagadevishal/why-schools-must-rethink-learning-in-the-age-of-ai-a4b478203888?source=rss-b56df248cb69------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@kagadevishal/why-schools-must-rethink-learning-in-the-age-of-ai-a4b478203888?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/a4b478203888</guid>
            <category><![CDATA[education]]></category>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[schools]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[nano-banana]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Mon, 01 Dec 2025 15:03:19 GMT</pubDate>
            <atom:updated>2025-12-01T15:39:22.741Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[How to Make LLMs Stop Agreeing with You All the Time. Is it even possible?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://pub.towardsai.net/how-to-make-llms-stop-agreeing-with-you-all-the-time-is-it-even-possible-22a32a527ef0?source=rss-b56df248cb69------2"><img src="https://cdn-images-1.medium.com/max/735/1*woqUhB6rZSZEBcohYQoAPQ.png" width="735"></a></p><p class="medium-feed-snippet">Are you also frustrated that your LLM never challenges your thoughts? It just agrees with whatever you say, which makes brainstorming or&#x2026;</p><p class="medium-feed-link"><a href="https://pub.towardsai.net/how-to-make-llms-stop-agreeing-with-you-all-the-time-is-it-even-possible-22a32a527ef0?source=rss-b56df248cb69------2">Continue reading on Towards AI »</a></p></div>]]></description>
            <link>https://pub.towardsai.net/how-to-make-llms-stop-agreeing-with-you-all-the-time-is-it-even-possible-22a32a527ef0?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/22a32a527ef0</guid>
            <category><![CDATA[sycophancy]]></category>
            <category><![CDATA[writing-prompts]]></category>
            <category><![CDATA[prompt-engineering]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[llm]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Wed, 26 Nov 2025 00:19:19 GMT</pubDate>
            <atom:updated>2025-11-27T02:00:01.784Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[The Only Image Generation Prompting Guide You Will Ever Need]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://pub.towardsai.net/the-only-image-generation-prompting-guide-you-will-ever-need-a7e06bbf0bbe?source=rss-b56df248cb69------2"><img src="https://cdn-images-1.medium.com/max/1239/1*UgWhUzcR_Z6PDLX8ckhaxw.png" width="1239"></a></p><p class="medium-feed-snippet">Non members can read story here.</p><p class="medium-feed-link"><a href="https://pub.towardsai.net/the-only-image-generation-prompting-guide-you-will-ever-need-a7e06bbf0bbe?source=rss-b56df248cb69------2">Continue reading on Towards AI »</a></p></div>]]></description>
            <link>https://pub.towardsai.net/the-only-image-generation-prompting-guide-you-will-ever-need-a7e06bbf0bbe?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/a7e06bbf0bbe</guid>
            <category><![CDATA[image-generation]]></category>
            <category><![CDATA[stable-diffusion]]></category>
            <category><![CDATA[prompt]]></category>
            <category><![CDATA[prompt-engineering]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Thu, 20 Nov 2025 12:55:21 GMT</pubDate>
            <atom:updated>2025-12-02T23:43:46.837Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Why TPUs Never Became the Next GPUs]]></title>
            <link>https://ai.plainenglish.io/why-tpus-never-became-the-next-gpus-7dc5692c5d93?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/7dc5692c5d93</guid>
            <category><![CDATA[google]]></category>
            <category><![CDATA[gpu]]></category>
            <category><![CDATA[tpu]]></category>
            <category><![CDATA[neural-networks]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Sun, 16 Nov 2025 09:44:10 GMT</pubDate>
            <atom:updated>2025-12-03T17:44:08.253Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Xt1wmiYExwJQbSB013SwMA.png" /><figcaption>created by AI using gemini nano banana</figcaption></figure><p>Edit : After the release of Gemini 3, Google will probably take the lead in the AI race, not because it has created a <strong>state of the art</strong> model but because it is trained entirely on TPUs using Google’s <strong>inhouse infrastructure</strong>. Their reduced reliance on GPUs may become a <strong>long term game changer</strong>.</p><p>If you are into machine learning, you have probably heard about Google’s TPUs. They are fast, they are efficient, and they are cheaper than many Nvidia GPUs. So why is the entire industry still obsessed with Nvidia rather than lining up for TPUs.</p><p>It turns out the answer is not one thing. It is a <strong>mix of history, accessibility, ecosystem, and good old business strategy</strong>.</p><p>Let us start with the simplest reason!</p><h4>You cannot buy them</h4><p>TPUs mostly live inside Google Cloud. They are not sold like GPUs. There is no TPU in your desktop and you will not find one on Amazon.Because,</p><blockquote>Most of them are not for sale.</blockquote><p>Google built TPUs for itself. They were never meant to be a consumer product or a hardware empire like Nvidia cards. The TPU project started because Jeff Dean (<em>Head of google AI that time</em>) ran some numbers in 2013 on napkin they say and realized that if every <strong>Android user </strong>used voice recognition for <strong>one minute a day</strong>, Google’s entire data center capacity <strong>would collapse</strong>. So Google built custom silicon to solve a very Google specific problem.</p><p>And it worked. So well that they never needed to sell it.</p><h4>GPUs got popular through gaming not AI</h4><p>Long before deep learning became mainstream,GPUs were powering video games. Gamers wanted faster graphics and companies like Nvidia built hardware for that market. GPUs became common in every desktop PC and later every laptop. Developers already knew how to program them. Then CUDA arrived and everything changed.</p><p>CUDA turned GPUs into the default platform for machine learning. You could hire any CS graduate and they probably knew CUDA or could learn it quickly. You could train models on the same GPU you played games on. You could install one at home or buy one in the cloud.</p><p>TPUs do not have that luxury. If you want to use them seriously you probably need to use Google Cloud. If you want to write optimized code you probably need JAX or TensorFlow XLA. And if you want to debug a kernel issue on a TPU you might need to hire a former Google engineer.</p><h4>Ecosystem beats raw power</h4><p>This is the probably is the <em>game-changer</em>.Hardware is only half the story. Nvidia spent years building an ecosystem. CUDA, cuDNN, PyTorch support, TensorRT, massive community documentation, endless tutorials and tons of open source tooling built around GPUs. Research labs prototype on GPUs. Startups deploy on GPUs. Even TPU compatible frameworks like JAX run perfectly well on Nvidia hardware.</p><p>TPUs have amazing performance for matrix multiplication. That is what they are born to do. But deep learning evolves fast. Sparse ops, low bit inference, MoE, custom kernels, new architectures. GPUs are flexible enough to keep up. TPUs are incredibly efficient but less general purpose.</p><h3>Another big factor they lock you in</h3><p>If you go all in on TPUs you go all in on Google Cloud. That is risky. With GPUs you can train on AWS, run experiments on a workstation at home, deploy to Azure, or run inference on a Jetson. You have options. Companies love options.</p><p>TPUs give you one option <strong>Google</strong>.</p><p><strong><em>So are TPUs dead Not at all.</em></strong></p><p>TPUs still power massive parts of Google. They help train Gemini. They are used by many AI startups on GCP. TPUs are extremely cost efficient for large dense training jobs.</p><p>Even your phone probably has something like a TPU inside it for on device AI. And Nvidia GPUs now include tensor cores which are basically TPU style blocks inside a GPU. The future might be a blend of both approaches.</p><blockquote><strong>So why the lack of hype</strong></blockquote><p>It is not because TPUs are bad.<br>It is because the world was not built around them.</p><p>GPUs had a fifteen year head start, a gaming market, a developer community, and a company that decided to make them the default platform for every new AI researcher. TPUs had Google and Google kept them to itself.</p><p>If Google sold TPU based PCI cards in 2017 and pushed developers to use them the conversation today might be very different.</p><p>Until then Nvidia remains the king not because they make the fastest chips but because they built the best ecosystem.</p><h3>References</h3><ul><li><strong>The Jeff Dean napkin story (TPU origin)</strong><br> Luca Ferretti, <em>“Tensor Processing Units: enabling the next generation of fast affordable AI”</em> (Codemotion, 2019)<br> <a href="https://www.codemotion.com/magazine/ai-ml/machine-learning/tensor-processing-units-enabling-the-next-generation-of-fast-affordable-ai/?utm_source=chatgpt.com">https://www.codemotion.com/magazine/ai-ml/machine-learning/tensor-processing-units-enabling-the-next-generation-of-fast-affordable-ai/</a></li><li><strong>Google’s official TPU architecture and hardware details</strong><br> Google Cloud TPU System Architecture<br> <a href="https://cloud.google.com/tpu/docs/system-architecture">https://cloud.google.com/tpu/docs/system-architecture</a></li><li><strong>Original TPU performance paper</strong><br> Norm Jouppi et al, <em>“In Datacenter Performance Analysis of a TPU”</em> (Google Research, 2017)<br> <a href="https://arxiv.org/abs/1704.04760">https://arxiv.org/abs/1704.04760</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7dc5692c5d93" width="1" height="1" alt=""><hr><p><a href="https://ai.plainenglish.io/why-tpus-never-became-the-next-gpus-7dc5692c5d93">Why TPUs Never Became the Next GPUs</a> was originally published in <a href="https://ai.plainenglish.io/">Artificial Intelligence in Plain English</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Verbalized Sampling: The Hidden Trick to Unlock Real Creativity in ChatGPT]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://pub.towardsai.net/verbalized-sampling-the-hidden-trick-to-unlock-real-creativity-in-chatgpt-72350154da47?source=rss-b56df248cb69------2"><img src="https://cdn-images-1.medium.com/max/660/1*e6EOEY1J_v7P4C_sQAnNiw.png" width="660"></a></p><p class="medium-feed-snippet">Have you ever noticed that AI models often give boring, predictable and less creative answers, even when there are lot of responses.</p><p class="medium-feed-link"><a href="https://pub.towardsai.net/verbalized-sampling-the-hidden-trick-to-unlock-real-creativity-in-chatgpt-72350154da47?source=rss-b56df248cb69------2">Continue reading on Towards AI »</a></p></div>]]></description>
            <link>https://pub.towardsai.net/verbalized-sampling-the-hidden-trick-to-unlock-real-creativity-in-chatgpt-72350154da47?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/72350154da47</guid>
            <category><![CDATA[writing-prompts]]></category>
            <category><![CDATA[prompt-engineering]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Fri, 14 Nov 2025 16:50:07 GMT</pubDate>
            <atom:updated>2025-11-17T04:10:45.171Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Why We need MCP (Model Context Protocol)?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://ai.plainenglish.io/why-we-need-mcp-model-context-protocol-0cf3f7338ebd?source=rss-b56df248cb69------2"><img src="https://cdn-images-1.medium.com/max/2600/1*v0oIH1BBjLop5A9kLisnuA.jpeg" width="5246"></a></p><p class="medium-feed-snippet">Introduction</p><p class="medium-feed-link"><a href="https://ai.plainenglish.io/why-we-need-mcp-model-context-protocol-0cf3f7338ebd?source=rss-b56df248cb69------2">Continue reading on Artificial Intelligence in Plain English »</a></p></div>]]></description>
            <link>https://ai.plainenglish.io/why-we-need-mcp-model-context-protocol-0cf3f7338ebd?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/0cf3f7338ebd</guid>
            <category><![CDATA[ai-infrastructure]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[mcp-protocol]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Mon, 10 Nov 2025 00:22:01 GMT</pubDate>
            <atom:updated>2025-11-17T12:33:52.429Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Extracting Files from Tarballs(tar) in Linux: A Comprehensive Guide]]></title>
            <link>https://medium.com/@kagadevishal/extracting-files-from-tarballs-tar-in-linux-a-comprehensive-guide-93dcd7e53818?source=rss-b56df248cb69------2</link>
            <guid isPermaLink="false">https://medium.com/p/93dcd7e53818</guid>
            <category><![CDATA[unzip]]></category>
            <category><![CDATA[large-file]]></category>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[zip]]></category>
            <category><![CDATA[extract]]></category>
            <dc:creator><![CDATA[Kagadevishal]]></dc:creator>
            <pubDate>Mon, 14 Oct 2024 10:53:52 GMT</pubDate>
            <atom:updated>2025-11-16T09:46:50.075Z</atom:updated>
            <cc:license>https://creativecommons.org/publicdomain/mark/1.0/</cc:license>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*O9QCzADF0u0Vld50ZtnxpA.png" /><figcaption>created by AI using gemini nano banana</figcaption></figure><p>In Linux, compressed archives are often packaged as <strong>tarballs</strong> (.tar, .tar.gz, .tgz, .tar.bz2). These files bundle multiple files and directories into a single file while optionally compressing them to save space. Extracting files from tarballs is a common task, and Linux provides powerful tools to handle this efficiently.</p><p>This guide will walk you through various methods to extract files from tarballs, including handling different compression formats and specific extraction options.</p><p><strong>Understanding Tarballs and Compression</strong></p><p>Before we dive into extracting, it’s important to understand the structure and compression of tarballs:</p><ul><li><strong>.tar</strong>: A simple archive format that bundles files without compressing them.</li><li><strong>.tar.gz or </strong><strong>.tgz</strong>: A tarball that is compressed using <strong>gzip</strong>.</li><li><strong>.tar.bz2</strong>: A tarball compressed using <strong>bzip2</strong>, which offers better compression than gzip but is slower.</li></ul><p><strong>Basic Extraction with </strong><strong>tar</strong></p><p>The tar command is the primary tool for working with tarballs. It can create, compress, and extract tar files with a variety of options.</p><h4>Syntax for Extracting Tarballs</h4><p>The basic syntax for extracting a tarball is:</p><pre>!tar -xvf archive_name.tar</pre><p>Here’s what each option means:</p><ul><li><strong>-x</strong>: Extract the files from the archive.</li><li><strong>-v</strong>: Verbose mode, which shows progress during the extraction (optional).</li><li><strong>-f</strong>: Specifies the tarball file.</li></ul><p><strong>Extracting Compressed Tarballs</strong></p><p>Depending on the compression type, you need to add an extra option:</p><ul><li><strong>For </strong><strong>.tar.gz or </strong><strong>.tgz</strong> (gzip compression), use the -z flag:</li></ul><pre>!tar -xzvf archive_name.tar.gz</pre><ul><li><strong>-z</strong>: Tells tar to use gzip for compression/decompression.</li><li><strong>For </strong><strong>.tar.bz2</strong> (bzip2 compression), use the -j flag:</li></ul><pre>tar -xjvf archive_name.tar.bz2</pre><p><strong>-j</strong>: Tells tar to use bzip2 for compression/decompression.</p><h3>Extracting Specific Files or Directories</h3><p>You don’t always need to extract everything from a tarball. To extract specific files or directories, specify their paths after the tarball name:</p><pre>tar -xzvf archive_name.tar.gz path/to/file</pre><p>This extracts only the specified file or directory from the tarball. You can list multiple paths for extraction:</p><pre>tar -xzvf archive_name.tar.gz path/to/file1 path/to/file2</pre><h3>Extracting Tarballs Without Overwriting Existing Files</h3><p>If you’re extracting files into a directory that already contains files, and you want to avoid overwriting existing files, you can use the --skip-old-files option:</p><pre>!tar -xzvf archive_name.tar.gz --skip-old-files</pre><p>This command will extract only the files that don’t already exist in thedestination directory, leaving existing files untouched.</p><h3>Handling Large Tarballs: Extracting Specific Parts</h3><p>For very large tarballs, you may want to extract only a portion of the archive to avoid memory or storage constraints.</p><ul><li><strong>Extract only a few files</strong>: Use the method shown above to extract specific files or directories.</li><li><strong>Extract from a specific directory in the tarball</strong>:</li></ul><pre>!tar -xzvf archive_name.tar.gz --strip-components=1</pre><ul><li>The --strip-components option removes a specified number of leading directory components from the paths in the archive. For example, if your archive contains files in a folder called archive_root, using --strip-components=1 will extract the files directly without the archive_root directory.</li></ul><p><strong>Extracting Tarballs with Pipes</strong></p><p>You can also extract tarballs directly from downloads or pipes without saving the tarball to disk. For instance, if you’re downloading a tarball from the internet, you can extract it as it’s being downloaded:</p><pre>!curl -O http://example.com/archive.tar.gz | tar -xzv</pre><p>This command downloads the file from the URL and extracts it without saving the tarball locally.</p><h3>Dealing with Permissions</h3><p>Sometimes, extracted files may have permission issues. You can preserve file permissions by default using the -p option:</p><pre>tar -xzvpf archive_name.tar.gz</pre><ul><li><strong>-p</strong>: Preserves the original file permissions from the archive.</li></ul><h3>Extracting Large or Corrupted Tarballs</h3><p>In some cases, a tarball might be very large or partially corrupted, making extraction difficult. Here are some tips for handling these cases:</p><ul><li><strong>Resume extraction</strong>: If the extraction process fails halfway, you can restart it by adding the --ignore-zeros option to skip over null blocks:</li></ul><pre>tar -xzvf archive_name.tar.gz --ignore-zeros</pre><ul><li><strong>Extracting large files</strong>: For large archives, ensure you have sufficient disk space. If you are extracting on a remote server with limited space, consider extracting specific portions of the archive first, as described above.</li></ul><p><strong>Extracting specific folder from the tarbell file</strong></p><p>Rather than extracting complete folder, if you only want to extract a certain folders from the tar file, you can do as follows:</p><pre>!tar -xfz archive_name.tar.gz -c destination_folder folder1 folder2</pre><p>so you can assign the folder you want to extract after the destination folder.</p><p>I hope this helps.</p><p>BYE!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=93dcd7e53818" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>