<?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 Khizar Shah on Medium]]></title>
        <description><![CDATA[Stories by Khizar Shah on Medium]]></description>
        <link>https://medium.com/@khizarshah01?source=rss-58044a6b4850------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*oz7fr0ZzCXrRJts4rqOc1Q.jpeg</url>
            <title>Stories by Khizar Shah on Medium</title>
            <link>https://medium.com/@khizarshah01?source=rss-58044a6b4850------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 13:26:48 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@khizarshah01/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[5 Claude Code Skills That Feel Like Cheating for Web Design (Most Vibe Coders Miss These)]]></title>
            <link>https://khizarshah01.medium.com/5-claude-code-skills-that-feel-like-cheating-for-web-design-most-vibe-coders-miss-these-a8e8cdc0284d?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/a8e8cdc0284d</guid>
            <category><![CDATA[web-design]]></category>
            <category><![CDATA[vibe-coding]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[claude]]></category>
            <category><![CDATA[programming]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Fri, 20 Nov 2026 11:11:09 GMT</pubDate>
            <atom:updated>2026-11-20T11:11:09.162Z</atom:updated>
            <content:encoded><![CDATA[<p>Typing “make it modern” doesn’t give your AI taste. These five free skills do: style, motion, scroll effects, 3D and a final polish. One command each.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vJsUsejQOTq5U0QAlOPQmQ.png" /></figure><p>You ask Claude for a landing page.</p><p>Thirty seconds later you get it. Purple gradient. A big centered headline. Three cards with tiny icons. A fade-in on everything.</p><p>You know this website. You’ve seen it a hundred times. If you’ve built with AI, you’ve probably shipped it too. Most of us have.</p><p>Here’s the thing. The AI isn’t bad at design. It just has no taste unless you give it some.</p><p>And typing “make it clean and modern, add smooth animations” into every new chat? That’s not taste. That’s a wish.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hCZ4QTNfQwpoqBFrn0HWhw.png" /><figcaption>Left: what “make it modern” gets you. Right: the kind of page these skills push you toward. Same app, same message. (Both are mockups I made for this post.)</figcaption></figure><p>A few days ago I found a carousel by <strong>@nocodealex</strong> on Instagram listing 5 Claude Code skills that make builds look <em>designed</em> instead of <em>default</em>. I stopped scrolling.</p><p>So I did the nerd thing. I opened all five repos, read every README, and checked how each one actually installs.</p><p>None of them needs you to be a designer. Each one below comes with a real-life example, a prompt you can copy, and one install command.</p><h3>The Short Answer</h3><p>Short on time? Here’s the whole post in five lines.</p><ul><li><strong>Design DNA</strong> copies the style of a site you love.</li><li><strong>Motion Design</strong> makes your animations feel natural, not random.</li><li><strong>GSAP</strong> gives you smooth animations that play as you scroll.</li><li><strong>Three.js</strong> helps Claude write 3D code that actually runs.</li><li><strong>Genjutsu</strong> does the final polish before you ship.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fUfOTJQ-dQgO63o-ga3tsg.png" /><figcaption>Screenshot this one for later.</figcaption></figure><h3>Wait, what’s a “skill”?</h3><p>Simple version: <strong>a prompt is something you type every time. A skill is something you install once.</strong></p><p>Think of it like cooking. A prompt is explaining your biryani recipe to a friend every single time. A skill is handing them the recipe book once.</p><p>Under the hood, a skill is just a folder with a file called SKILL.md. It has a short line at the top saying when to use it, and instructions below.</p><p>Claude Code only opens a skill when your request matches it. The rest of the time, it just sits there quietly.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5CrtKCAcApN2j9yJ0yUOhg.jpeg" /><figcaption>Me after finding out skills exist.</figcaption></figure><p>Most of these install with one command: npx skills add .... You just need Node.js on your computer. Okay, let&#39;s go.</p><h3>1. Design DNA: steal a vibe (legally)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NE9MovaHcVcPl-A2t7a_WA.png" /></figure><p><strong>In one line:</strong> you show it a website you love, and it builds your page in that same style.</p><p>This is my favorite idea on the list, so it goes first.</p><p>Say you’re building your portfolio. You love how some site looks, but “make it look like that” never works. Claude kind of gets it, then drifts back to purple gradients.</p><p>Design DNA fixes that. It studies the site (a link or screenshots) and writes down its colors, fonts, spacing and mood in one file. Then it uses that file to build <em>your</em> content.</p><p>You can save that file and reuse it on your next project too. Your style, written down once.</p><p>My favorite detail: AI is bad at seeing exact colors. It sees a brand pink and guesses a more common pink. Design DNA comes with small scripts that measure the real colors from the image, then check if the result matches.</p><p><strong>Copy this prompt:</strong></p><blockquote><em>“Here’s a site I love: linear.app. Extract its Design DNA, then build my portfolio homepage in that style.”</em></blockquote><p><strong>Install:</strong></p><pre>npx skills add zanwei/design-dna</pre><p><strong>Repo:</strong> github.com/zanwei/design-dna (~2.1k stars)</p><h3>2. Motion Design: animations that feel right</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JZqlKhn9zNI85vbI4gtspQ.png" /></figure><p><strong>In one line:</strong> it teaches Claude <em>when</em> and <em>why</em> things should move, not just how.</p><p>Be honest. The biggest giveaway of an AI site isn’t <em>no</em> animation. It’s <em>the same</em> fade-in on every single thing.</p><p>This skill comes from LottieFiles, a big name in web animation. It teaches your AI to think like a motion director. How long something should take. Whether it should start fast or slow. What moves first and what follows.</p><p>It even adapts Disney’s 12 classic animation principles for buttons and cards. Yes, the cartoon ones.</p><p>Say you have a “Submit” button. Right now it just… submits. With this skill, Claude can make it show a loading state, then a little success moment, or a clear error if something breaks.</p><p>It works with whatever you already use: plain CSS, Framer Motion, GSAP, Lottie and more.</p><p><strong>Copy this prompt</strong> (straight from their README):</p><blockquote><em>“Create a loading → success → error state transition for this form submit button.”</em></blockquote><p><strong>Install:</strong></p><pre>npx skills add LottieFiles/motion-design-skill</pre><p><strong>Repo:</strong> github.com/LottieFiles/motion-design-skill (~3.4k stars)</p><h3>3. GSAP: smooth scroll magic</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PXp06Z8aSLvLGXkuJQPtUg.png" /></figure><p><strong>In one line:</strong> the official skills from the team behind GSAP, the animation library behind a lot of those smooth websites you screenshot for “inspiration”.</p><p>Not a random fan repo. The <em>official</em> one, and the most starred on this list.</p><p>Say you want your features section to stay pinned on screen while each feature slides in, one by one, as you scroll. GSAP has a tool for exactly that, called ScrollTrigger. It ties animations to your scroll position.</p><p>The skill teaches Claude to use it properly, plus timelines (animations that play in order instead of all at once). It covers React, Vue, Svelte and plain JavaScript.</p><p>Fun fact from the README: GSAP is now 100% free, including the plugins that used to be paid, like SplitText.</p><p><strong>Copy this prompt:</strong></p><blockquote><em>“Use GSAP ScrollTrigger to pin the features section and reveal each feature one by one as I scroll.”</em></blockquote><p><strong>Install:</strong></p><pre>npx skills add https://github.com/greensock/gsap-skills</pre><p>Or, inside Claude Code: /plugin marketplace add greensock/gsap-skills</p><p><strong>Heads up:</strong> if your project already uses a different animation library, the skill tells Claude to respect that. Nice touch.</p><p><strong>Repo:</strong> github.com/greensock/gsap-skills (~18.8k stars)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/857/1*O3xTw3wlzNrVweS1qHFGIw.jpeg" /><figcaption>The four stages of AI web design.</figcaption></figure><h3>4. Three.js: 3D that actually works</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zA-2-fbeFX7PuworBq0tiA.png" /></figure><p><strong>In one line:</strong> it teaches Claude a recent version of three.js, the most popular 3D library for the web.</p><p>Here’s the problem. Claude learned three.js from years of old tutorials, so it mixes old code with new code. Some of those old names don’t exist anymore, and that’s exactly what breaks your 3D scene.</p><p>This repo gives Claude 13 small skills, one per topic: lighting, materials, loading 3D models, performance and more. Each one ends with a “Common Mistakes” list of old patterns to avoid.</p><p>Say you’re making a product page and want the product to slowly spin in the hero. Or a background that gently moves. That’s where this one shines.</p><p><strong>Copy this prompt:</strong></p><blockquote><em>“Add a slowly rotating 3D object to the hero section with soft lighting, and keep it smooth on mobile.”</em></blockquote><p><strong>Install:</strong></p><pre>npx skills add cloudai-x/threejs-skills</pre><p><strong>Heads up:</strong> 3D is heavy. Use it where it actually adds something, not on every page.</p><p><strong>Repo:</strong> github.com/CloudAI-X/threejs-skills (~4.3k stars)</p><h3>5. Genjutsu: the final polish pass</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qIot5r5H8aaz_VK0wRUfEA.png" /></figure><p>Yes, it’s named after the Naruto illusion jutsu. The tagline is literally “The art of illusion.” I was sold before reading anything else.</p><p><strong>In one line:</strong> it plans how your page should look and move, waits for your okay, then builds it.</p><p>It’s actually three skills in one:</p><ul><li><strong>cast</strong> makes one part of your page move (a gallery, a navbar, a button)</li><li><strong>paint</strong> designs a product’s whole look</li><li><strong>bunshin</strong> builds a whole website with a team of AI helpers (yes, shadow clones)</li></ul><p>The best part? Before writing any code, it pitches its idea for how things should move and feel, and <strong>waits for your yes.</strong></p><p>It also keeps a list of the lazy defaults AI loves, like three equal cards in a row, so it can avoid them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*fXP8mbFhd14T1FQ9EXipig.gif" /><figcaption>“Nocturne”, a planetarium landing page from one of Genjutsu’s recorded example runs. Source: AThevon/genjutsu</figcaption></figure><p>Say your page works, but it feels flat. Your pricing cards just sit there. This is the skill for that last bit of shine.</p><p><strong>Copy this prompt</strong> (from the README):</p><blockquote><em>“Make the pricing cards feel physical on hover.”</em></blockquote><p><strong>Install:</strong></p><pre>npx skills add https://genjutsu.athevon.dev -g</pre><p>Use that link, not the repo name. The README says installing from the repo name skips some of its parts. You can also add it as a plugin inside Claude Code: /plugin marketplace add AThevon/genjutsu, then /plugin install genjutsu.</p><p><strong>Heads up:</strong> I love how honest the README is. It says it’s only been tested on Claude Code, and that in its own test a run cost about four times a normal run. Worth knowing before you go wild.</p><p><strong>Repo:</strong> github.com/AThevon/genjutsu (~820 stars, the smallest here, but actively updated)</p><h3>How they stack together</h3><p>Here’s the part that clicked for me. These aren’t five competing tools. They’re layers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Mg8XffCgnmrCnWYxnCgnuQ.png" /></figure><p>First, <strong>Design DNA</strong> picks the look. Then <strong>Motion Design and GSAP</strong> make it move. <strong>Three.js</strong> adds 3D, only if it helps. And <strong>Genjutsu</strong> polishes it before you ship.</p><p>Install them once, and they’re there in every project.</p><p>Two quick warnings. First, don’t throw every skill at every page. A simple blog doesn’t need a 3D scene and a pinned scroll animation.</p><p>Second, skills are instructions written by strangers. Open the SKILL.md and skim it before you install. It takes two minutes.</p><h3>Where should you start?</h3><p>If you only install one today, make it <strong>Design DNA.</strong> That’s the one I’m starting with.</p><p>Why? Because the look comes before the movement. A beautiful animation on a boring layout is still a boring layout, just moving.</p><p>Then add <strong>Motion Design</strong>, so your animations stop feeling random. Grab GSAP when you want scroll effects, Three.js when you want 3D, and Genjutsu for the final shine.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZiyhtAfgA68lAuw0m6D3aw.png" /><figcaption>Are ya winning, son? Finally, Dad. Finally.</figcaption></figure><h3>That’s it</h3><p>All five repos are public and free. Setup takes about five minutes. Your next project doesn’t have to look like everyone else’s.</p><p>I’m a student from a tier-3 college, selected for GSoC 2026 with Rocket.Chat, and I’m building in public. If this helped, <strong>follow me here on Medium</strong> for more posts like this. I’m also on X at <strong>@0xrinx</strong>, come say hi.</p><p>And if you’ve found a design skill better than these five, drop it in the comments. I’ll happily steal it. 😄</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a8e8cdc0284d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[My GSoC 2026 Journey: Building the Background Jobs Admin Page for Rocket.Chat]]></title>
            <link>https://khizarshah01.medium.com/my-gsoc-2026-journey-building-the-background-jobs-admin-page-for-rocket-chat-8a742e11ba46?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/8a742e11ba46</guid>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[gsoc-2026]]></category>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[google-summer-of-code]]></category>
            <category><![CDATA[rocket-chat]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Sun, 23 Aug 2026 08:07:29 GMT</pubDate>
            <atom:updated>2026-08-23T08:07:29.049Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HXoN6AG61xmphe9YnF-B0g.jpeg" /></figure><p>Hello! I am Khizar Shah, and over the past few months, I had the incredible opportunity to participate in Google Summer of Code (GSoC) 2026 with Rocket.Chat. My project was to build a comprehensive <strong>Background Jobs Administration Panel</strong> that giving workspace administrators total visibility and control over all Agenda-based background jobs.</p><p><strong>Quick links:<br></strong><a href="https://github.com/RocketChat/Rocket.Chat/pull/40699">Pull Request</a><br><a href="https://github.com/Khizarshah01/GSoC-26">Report</a><br><a href="https://summerofcode.withgoogle.com/programs/2026/projects/fPHB4nLt">GSoC’26 Official page</a><br>Demo Video:</p><iframe src="https://cdn.embedly.com/widgets/media.html?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D5KMxUgt5cVk&amp;type=text%2Fhtml&amp;schema=youtube&amp;display_name=YouTube&amp;src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F5KMxUgt5cVk" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/a0891cfd354c77cf13ba1574adb42f3a/href">https://medium.com/media/a0891cfd354c77cf13ba1574adb42f3a/href</a></iframe><h4><strong>The Problem</strong></h4><p>Rocket.Chat is a massive platform that relies on background jobs for everything from sending emails to auto-closing inactive live chats etc. It uses <a href="https://github.com/agenda/agenda">Agenda</a> as its scheduling engine.</p><p>Before this project, administrators had a major blind spot. they didn’t know which jobs existed, whether they were failing, or when they last ran. To make matters more complex, these jobs didn’t live in one place in database. They were fragmented across multiple Agenda collections (system cron, third-party apps, and omnichannel schedulers).</p><p>Without a single place to search, filter, inspect history, or manually trigger a system job, debugging scheduled tasks was a frustrating experience for support teams and server admins.</p><h4>What I Shipped</h4><p>The final deliverable is a fully functional Admin page integrated into the Rocket.Chat workspace. Here is a high-level look at what was built:</p><ul><li><strong>Admin UI:</strong> A dedicated page at Admin → Background Jobs featuring a clean 3-tab layout (System, Apps, Omnichannel).</li><li><strong>Actions:</strong> Safe, manual controls to <strong>Run Now</strong>, <strong>Enable</strong>, and <strong>Disable</strong> only for system jobs.</li><li><strong>UX:</strong> Server-side search, in-memory status filtering, and a contextual bar for viewing job details and execution history.</li><li><strong>API:</strong> 8 new REST endpoints (e.g., cron.jobs, cron.trigger) protected by a new manage-scheduled-jobs permission.</li><li><strong>History Logging:</strong> A shared CronHistory collection that securely logs the start, success, and error traces of system, app, and omnichannel jobs.</li></ul><h4>Architecture</h4><p>The admin page does not talk to MongoDB or Agenda collections directly.<br>Requests flow through a small stack:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*i2qP88EK5B-W7NHmCy3RUg.png" /><figcaption>simple architecture of the Background Jobs admin feature</figcaption></figure><p>1. The <strong>Admin UI</strong> calls the <strong>REST APIs</strong> (<em>cron.*</em>), gated by manage-scheduled-jobs.<br>2. <strong>CronJobsService</strong> lists jobs, resolves status, reads history, and handles system actions (enable / disable / run now).<br>3. Job data lives in <strong>System / Apps / Omnichannel</strong> collections. <strong>CronHistory</strong> stores execution logs.<br>4. When a job actually runs, the schedulers update status and write history (shown as the dotted path on the diagram).</p><h4>Deep Dives</h4><p><strong>A) The Job Status Dilemma<br></strong>Agenda does not natively provide a perfect, human-readable “status” string for admins. It relies on raw timestamps (lockedAt, nextRunAt, failedAt).</p><p>My first instinct was to simply save a status string to the database. But what happens if the server crashes while a job is running? The job&#39;s lock expires, but the database still says &quot;running&quot;.</p><p>To solve this, we implemented a <strong>Denormalized + Derived</strong> architecture:</p><ol><li>We save a status field and update it on standard lifecycle events (create, run, fail, enable).</li><li>If the status is missing, or if it is stuck on &quot;running&quot; with an expired lock, the API passes the job through a deriveStatus() algorithm. This algorithm recomputes the absolute truth based on lockedAt, fail counts, and nextRunAt.</li><li>Because the final status is resolved on the fly, we perform status filtering (e.g., viewing only “failed” jobs) <em>in memory</em> after the query. Given that a workspace typically runs a few dozen jobs, this trade-off heavily prioritized accuracy over raw database query speed.</li></ol><p><strong>B) Safe “Run Now” (Avoiding Duplicates)<br></strong>Implementing the “Run Now” feature seemed simple: just call agenda.now(). However, I discovered that agenda.now() creates an entirely new, duplicate job row in the database rather than executing the existing job definition.</p><p>To prevent polluting the database, I changed the approach. Instead of forcing a new job, the backend fetches the existing job, overwrites its nextRunAt property to the current timestamp (new Date()), and saves it. The Agenda engine picks it up immediately, executing the existing job without creating duplicates.</p><h4>Other Challenges</h4><ul><li><strong>Monorepo Complexities:</strong> Rocket.Chat is a massive monorepo. Learning to navigate packages, use the service broker, and adhere to strict CI linting habits took time but vastly improved my code quality.</li><li><strong>Omnichannel Realities:</strong> Omnichannel jobs (like auto-closing a chat) are often one-time, per-room jobs. They appear as “scheduled” while a chat is on hold, which initially felt like a bug but is actually the intended product behavior. Handling these transient jobs required special UX considerations.</li></ul><h4><strong>Scope: Shipped vs. Extras</strong></h4><p>A critical lesson I learned was project scope management. As we built the feature, we identified one “extra” tasks:</p><ol><li><strong>Orphan App Jobs:</strong> Investigating why jobs remain in the database after a third-party app is uninstalled.</li></ol><p>Following my mentors’ advice, I explicitly moved these out of scope for the final GSoC submission. Instead of spreading myself too thin, I prioritized polishing the core UI, the status derivation engine, and the API documentation to ensure a high-quality, stable final delivery.</p><h4>What I Learned</h4><p>Beyond shipping features, this project changed how I approach real production systems:</p><ul><li><strong>Building on systems you don’t fully control</strong> Agenda stores raw timestamps (lockedAt, nextRunAt) but doesn&#39;t provide a simple &quot;status&quot; field that an admin dashboard needs. I learned to combine a stored status field (updated on lifecycle events) with deriveStatus / resolveStatus so the UI stays accurate even when status is missing or stuck as &quot;running&quot; after a crash.</li><li><strong>Verify library helpers against real side effects</strong> “Run Now” via agenda.now() looked correct until it created duplicate job documents. That taught me to check what a helper actually writes to the database, not only whether the job appears to run.</li><li><strong>Scope management is part of engineering</strong> With my mentors, we treated orphan app jobs as extras and focused on polishing the core admin panel, APIs, history, and status model. A strong GSoC outcome is a stable, reviewable deliverable, not every follow-up idea in the same PR.</li><li><strong>Working in a large monorepo</strong> Rocket.Chat spans many packages. I got better at placing code in the right layer (UI → REST → service → collections), running lint/typecheck on packages I touch, and treating CI as part of the craft.</li></ul><p>Overall, GSoC with Rocket.Chat pushed me from “make it work” toward “make it correct, reviewable, and honest about limits.”</p><p>I want to extend a massive thank you to my mentors, <strong>Douglas Gubert</strong> and <strong>Kevin Aleman</strong>. Their technical guidance, patience, and push for strict architectural separation made this project a success and significantly leveled up my engineering skills.</p><p>Thank you to the Rocket.Chat community and Google Summer of Code for an unforgettable summer!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8a742e11ba46" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Posting My Thoughts Online Earned Me $400+]]></title>
            <link>https://khizarshah01.medium.com/how-posting-my-thoughts-online-earned-me-400-4c016422f163?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/4c016422f163</guid>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[activated-thinker]]></category>
            <category><![CDATA[personal-growth]]></category>
            <category><![CDATA[social-media]]></category>
            <category><![CDATA[self-improvement]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Thu, 11 Jun 2026 17:50:14 GMT</pubDate>
            <atom:updated>2026-06-11T17:50:14.244Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Sometimes, the things you post in 30 seconds can create opportunities worth hundreds of dollars.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/659/1*4uN-KfdR7meui0uY7XWf9g.png" /></figure><p>A year ago, if someone told me I’d earn over $400 just by sharing my thoughts online, I would’ve laughed.</p><p>I wasn’t selling a course. <br>I wasn’t running ads.<br>I didn’t have a massive audience. <br>I was just posting.</p><p>Thoughts. Experiences. Things I was learning. Small wins. Occasional failures.</p><p>That’s it.</p><p>Then something interesting happened.</p><p>People started noticing.</p><p>A tweet about open source got traction. Another one about building projects reached people I didn’t know. Some posts about GSoC, coding, and learning in public started getting shared.</p><p>Here is one of them</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/658/1*oPeThXyA4No2NvcM2qB9pw.png" /></figure><p>At first, it was just likes.</p><p>Then came followers.<br>Then came DMs.</p><p>And eventually, opportunities.</p><p>Someone wanted me to promote their product.<br>Someone wanted feedback on their startup.<br>Someone wanted a retweet. Someone wanted access to my audience.</p><p><em>That’s when I realized something:</em></p><p><strong>Attention is a currency.</strong></p><p>Not the fake attention.<br>Not engagement-bait attention.</p><p>Real attention.</p><p>The kind you earn by showing up consistently and sharing useful things.</p><p>The funny part?<br>Most of my posts take less than two minutes to write.</p><p>No complicated content strategy.<br>No growth hacks.<br>Just thoughts.</p><p>Here’s my analytics from the last 4 weeks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qEujxsCnmCSit97DjMgenw.png" /></figure><p>You’ll notice there isn’t a single post with 1M, 2M, or 8M views.</p><p>Most posts received between 100K and 600K impressions.</p><p>That’s the point.</p><p>You don’t need to go viral to create opportunities. Consistent visibility is often enough.</p><p>Here’s what helped me:</p><h3>Share what you’re learning</h3><p>You don’t need to be an expert.</p><p>People connect more with someone who’s one step ahead than someone who’s pretending to know everything.</p><p>I posted about <strong>open source contributions</strong>, coding projects, GSoC preparation, and lessons from mistakes.</p><p>People followed the journey.</p><h3>Document, don’t manufacture</h3><p>Many people sit down thinking:</p><p>“What content should I create today?”</p><p>Instead, ask:</p><p>“<strong>What did I learn today?</strong>”</p><p>The second question is easier and more authentic.</p><h3>Opportunities follow visibility</h3><p>The internet can’t reward work it can’t see.</p><p>You might be building amazing things, but if nobody knows, nobody can help, hire, recommend, or collaborate with you.</p><p>Posting creates surface area for luck.</p><h3>Consistency beats viral moments</h3><p>Most of my opportunities didn’t come from a single viral post.</p><p>They came from months of showing up.<br>One post gets noticed.</p><p>Another builds trust.<br>A third creates familiarity.</p><p>Over time, people remember your name.</p><h3>Be a person, not a brand</h3><p>The posts that perform best are usually the most human ones.</p><p>Not polished.</p><p>Not corporate.</p><p>Just honest observations.</p><p>People follow people.</p><p>Not logos.</p><p>Not personal brands.</p><p>People.</p><p>Looking back, the $400 wasn’t the biggest reward.</p><p>The bigger reward was realizing that sharing my thoughts created opportunities I would’ve never found otherwise.</p><p>New connections.</p><p>New friends.</p><p>New projects.</p><p>New doors opening.</p><p>And 3000+ followers.</p><p>All because I clicked “Post.”</p><p>If you’re building something right now, don’t wait until you’re an expert.</p><p>Share what you’re learning.<br>Share what you’re struggling with.<br>Share what you’re building.</p><p>The internet rewards people who show up consistently.</p><p>You never know which post will change everything.</p><p>The first few posts will probably get ignored. Mine did too. Keep posting anyway.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4c016422f163" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I got selected for GSoC 2026 from a Tier-3 college]]></title>
            <link>https://khizarshah01.medium.com/how-i-got-selected-for-gsoc-2026-from-a-tier-3-college-5f1bab05f3c3?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/5f1bab05f3c3</guid>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[google-summer-of-code]]></category>
            <category><![CDATA[rocket-chat]]></category>
            <category><![CDATA[gsoc-2026]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Mon, 08 Jun 2026 13:33:08 GMT</pubDate>
            <atom:updated>2026-06-08T13:33:08.674Z</atom:updated>
            <content:encoded><![CDATA[<p>I come from a <strong>tier-3 college</strong> in Maharashtra, where most people around me had never even heard of Google Summer of Code.</p><p>I <strong>wasn’t a topper</strong>. My CGPA was around <strong>7.7(CGPA)</strong>. I had <strong>no internship</strong> experience, no big-tech connections.</p><p>Yet, in 2026, my name appeared on the list of students selected for <strong>Google Summer of Code</strong> with <strong>Rocket.Chat</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*blqiyhF0H52o5HKKwMwMeA.png" /><figcaption><a href="https://summerofcode.withgoogle.com/programs/2026/organizations/rocketchat">https://summerofcode.withgoogle.com/programs/2026/organizations/rocketchat</a></figcaption></figure><p>This is the journey of how a student from a tier-3 college, an average academic record, and zero special advantages managed to get there.</p><p>My name is <strong>Khizar Shah</strong>, and I’m a Computer Science student at SSGMCE Shegaon.</p><p>Like many engineering students, I started my journey by learning web development and solving DSA problems in leetcode. Over time, I built projects with React, Next.js, Node.js, and MongoDB while constantly looking for ways to improve as a developer.</p><p><strong><em>But there was one thing missing.</em></strong></p><p>I had never worked on a large real-world codebase. When you’re building your own projects, you know exactly how everything works because you wrote it. You know where the bug is likely hiding. You know which file to open and what change might fix it.</p><p>A production codebase is completely different.</p><p>The first time I opened a large open-source project, I felt like I knew nothing.</p><p>Thousands of files. Years of development. Hundreds of contributors. Features connected to other features that I didn’t even know existed.</p><p>A simple bug that looked like a 5-minute fix could take hours just to understand.</p><p>I would open one file, which led to another file, which led to another function, and suddenly I had ten tabs open trying to figure out what was happening.</p><p>That was the moment I realized that building projects and understanding large-scale software are two very different skills. And that’s exactly why I became interested in open source.</p><h3>I Discovering GSoC</h3><p>Around that time, I started exploring open source more seriously.</p><p>While reading about open-source programs and contributor experiences, I thought, Why can&#39;t I try for GSoC. The idea sounded simple:<strong> contribute to open source, work with mentors, and get paid to build real software</strong>.</p><p>But after looking deeper, it felt intimidating.</p><p><strong>Thousands of students applied every year</strong>. Many already had<strong> years of open-source experience</strong>. Some were from <strong>top colleges</strong> and had i<strong>mpressive portfolios.</strong></p><p>For a while, I treated GSoC as something that was out of reach for me.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/631/1*wjfXzmiK46xqEuDgxV2-Sg.jpeg" /></figure><p>Still, the stories of previous contributors kept pulling me back.</p><p>The more I researched, the more I noticed something interesting. Most successful contributors didn’t start by writing proposals. <strong>They started by becoming contributors first.</strong></p><p>That changed my entire approach. Instead of asking, “How do I get selected for GSoC?”</p><p>I started asking, “<strong>How do I become useful to an open-source project?</strong>”</p><p>That question eventually led me to Rocket.Chat.</p><h4>Choosing an Organization</h4><p>For those unfamiliar with Google Summer of Code, the process is fairly straightforward.</p><blockquote>You choose an organization, start contributing to its open-source project, interact with the community, and eventually submit a project proposal during the application period. After that, all that’s left is to wait for the results.</blockquote><p>Of course, doing it successfully is much harder than it sounds.</p><p>One of the biggest decisions is choosing the right organization. While exploring different organizations, Rocket.Chat immediately stood out to me.</p><ul><li>First, it was built using technologies I was already interested in, such as TypeScript, React, Node.js, and MongoDB.</li><li>Second, the community was incredibly active and welcoming to new contributors.</li><li>Another reason was consistency. Rocket.Chat had been participating in Google Summer of Code for more than nine years. While there was no guarantee they would be selected again, their long history gave me confidence that they would likely continue participating in future editions.</li></ul><p><strong>This was important because I wanted to start contributing before the official GSoC organization list was announced.</strong></p><p>Even if GSoC didn’t work out, I would still gain valuable open-source experience. And if Rocket.Chat was selected again, I would already be familiar with the codebase and community instead of starting from scratch.</p><h4>My First Contributions</h4><p>Choosing Rocket.Chat was the easy part.</p><p>Actually contributing was much harder.</p><p>The first challenge wasn’t fixing bugs. It was setting up the project itself.</p><p><strong>Rocket.Chat is a large production</strong> application with years of development behind it. Getting everything running locally took time, and understanding how different parts of the system worked took even longer.</p><p>Once I finally got the project running, I started looking for issues that were suitable for new contributors.</p><p>I quickly realized that reading an issue and solving it were two completely different things.</p><p>A bug that looked simple on GitHub could require hours of investigation. I would trace function calls, read old pull requests, and jump between dozens of files just to understand what was happening.</p><p>Many times I thought I had found the solution, only to realize I had misunderstood the problem entirely.</p><p>Still, I kept showing up. Every bug taught me something.</p><p>Every review comment taught me something.</p><p>And slowly, the codebase started feeling a little less intimidating.</p><p>One moment that really motivated me was receiving a message from a Rocket.Chat community member thanking me for my contributions and wishing me luck for GSoC.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/886/1*Brk85a7PAFTN28O6b8hcSA.png" /></figure><p>It might seem like a small thing, but when you’re contributing to a project with hundreds of contributors, knowing that your work is being noticed gives you a huge confidence boost.</p><p>That message made me feel like I was no longer just exploring the project, I was becoming part of the community.</p><p>Before the GSoC application period even started, I had already become one of the top contributors on the Rocket.Chat leaderboard.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/983/1*hsDRuDI-Ppk3nkIOcRbuwA.png" /></figure><p>Before the application period began, I had already merged 11 pull requests into Rocket.Chat.</p><p>Some of these contributions were <strong>more than beginner-friendly fixes</strong>. They <strong>involved performance improvements, bug fixes</strong>, and production code that would eventually be shipped to Rocket.Chat users worldwide.</p><p>Seeing my code become part of an actual product used by thousands of people was a completely different feeling from building personal projects.</p><p>For the first time, I wasn’t just writing code for myself.</p><p>I was writing code that other people would use.</p><p>That experience taught me more about software engineering than any tutorial or course ever could.</p><h3>Writing the Proposal</h3><p>By the time the GSoC application period arrived, I already had months of contributions behind me.</p><p>This gave me a huge advantage.</p><p>Instead of spending my time learning the codebase from scratch, I could focus on designing a project that would create real value for Rocket.Chat.</p><p>I chose to work on <strong>Agenda Jobs Admin Page</strong>. But I didn’t want my proposal to be just a document full of ideas.</p><p>I wanted to prove that the idea was technically possible.</p><p>So before submitting my proposal, I built a Proof of Concept (POC).</p><p>This helped me identify challenges early, validate my approach, and receive better feedback from mentors. The proposal went through multiple revisions. Every review improved it.</p><p>Every discussion revealed something I had overlooked.</p><p>By the time I finally submitted it, I knew it was far stronger than the first version I had written.</p><h4>Selection Day</h4><p>selected for GSoC’2026 Rocket chat</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/386/1*F2-j1qMFhgfVWIAq_34FCg.gif" /></figure><h3>Tips for Future GSoC Aspirants</h3><ul><li><strong>Don’t rush to create PRs.</strong> Spend time understanding the codebase first. A few meaningful contributions are better than many low-quality ones.</li><li><strong>Help other contributors.</strong> Answer questions, help with setup issues, and support newcomers. Open source is more than just writing code.</li><li><strong>Start early.</strong> Everyone says this, but it genuinely matters. Contributors who are already active before the organization list is announced have a huge advantage.</li><li><strong>GSoC isn’t only about code.</strong> I’ve seen people get selected with fewer code contributions because they communicated well, participated in discussions, and understood the project deeply.</li><li><strong>Talk to mentors.</strong> Join meetings, ask questions, discuss ideas, and stay visible. Mentors are ultimately choosing the person they’ll work with for months.</li><li><strong>Focus on the mentor after choosing a project idea.</strong> A great proposal matters, but mentor confidence in you matters just as much.</li><li><strong>Build something before writing the proposal.</strong> A Proof of Concept can make your proposal far stronger than pages of theoretical ideas.</li><li><strong>Luck exists.</strong> yep. Sometimes your project receives fewer slots than expected, or multiple strong candidates apply for the same idea. You can’t control that part.</li><li><strong>Nothing is wasted.</strong> Even if you don’t get selected, you’ll gain experience, connections, and real-world engineering skills that stay with you long after GSoC ends.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Fh5oOete-ERXvE-Cp8keMg.png" /></figure><h3>Final Thoughts</h3><p>If you’re from a tier-3 college, don’t let that become an excuse.</p><p>Open source doesn’t care about your college name, your CGPA, or your background.</p><p>It cares about whether you can learn, contribute, communicate, and keep showing up consistently.</p><p>A year ago, GSoC felt impossible to me.</p><p>Today, I’m a GSoC 2026 contributor at Rocket.Chat.</p><p>If I can do it, there’s no reason you can’t start your own journey today.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5f1bab05f3c3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Understanding Rocket.Chat Monorepo for Beginners]]></title>
            <link>https://khizarshah01.medium.com/understanding-rocket-chat-monorepo-for-beginners-86f5f72baf9a?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/86f5f72baf9a</guid>
            <category><![CDATA[open-source]]></category>
            <category><![CDATA[beginner-developers]]></category>
            <category><![CDATA[gsoc]]></category>
            <category><![CDATA[rocket-chat]]></category>
            <category><![CDATA[monorepo]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Sat, 14 Feb 2026 12:51:38 GMT</pubDate>
            <atom:updated>2026-02-14T12:51:38.616Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>The official docs are always the best place to start.</strong><br>But if you want a simple explanation from a beginner’s point of view, this is for you.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DJY1MflDCnx3wPpsoyC1kA.jpeg" /><figcaption>Rocket.Chat</figcaption></figure><p>When I first cloned the <a href="https://github.com/RocketChat/Rocket.Chat">Rocket.Chat</a> repository, I was stuck.<br>The repo is massive. I didn’t know where to begin.</p><p>If you are also stuck, this blog will help you make sense of the structure without wasting hours trying to understand manually.</p><h4>First, What Is a Monorepo?</h4><p>Rocket.Chat uses a <strong>monorepo</strong>.</p><p>A monorepo means multiple projects and packages live inside a single repository. Instead of separate repos for frontend, backend… everything is organised in one place.</p><p>And that makes it a little hard to understand in the first place. But once you understand the structure, it actually makes sense.</p><h4>Structure of Rocket.Chat</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LIH4hZGVYNTMs0mGhVepNA.png" /></figure><p>Let’s focus only on the important parts.</p><p>apps/: This is one of the most important folders.<br>meteor/ : This is the main Rocket.Chat application. Both frontend and backend live here.</p><p>If you are contributing to features or fixing bugs, most of your time will be inside apps/meteor/.</p><p>packages/<br>This folder contains many shared packages, reusable components used across the apps<em>. <br></em>I have explored some packages. Here are a few that helped me understand:</p><p>1.packages/message-parser : Contains all markdown parsing logic (<strong>bold</strong>, <em>italic</em>, etc…).</p><p>2.packages/i18n : Handles translations so Rocket.Chat can support multiple languages.</p><p>…. There are a lot of packages there, and if you start to read one by one, it is hard to understand all of them at once, but the best way is to <strong>only explore what is related to your issue.</strong></p><h4>Understanding The Codebase: “<strong>Fuselage</strong>” vs “<strong>UI Kit</strong>”</h4><p>This part is often confusing because <strong>Fuselage</strong> is a separate library, while the others are local packages. Let’s clear this up:</p><ol><li><strong>Fuselage (</strong><a href="https://github.com/RocketChat/fuselage"><strong>Rocket.chat/fuselage</strong></a><strong>)</strong>: This is an <strong>External Library</strong>.</li></ol><ul><li><strong>Where is it?</strong> It is NOT in the packages/ folder. It lives in its own repository (<a href="https://github.com/RocketChat/fuselage">RocketChat/fuselage</a>) and is installed as a dependency (just like React).</li><li><strong>What is it?</strong> It is the <strong>Source of Truth</strong> for design (Buttons, Colors, Fonts). Think of it like <strong>Bootstrap, </strong>a design system that provides reusable UI components.</li></ul><p><strong>2. UI Kit (</strong><strong>packages/ui-kit)</strong>: This is a <strong>Local Package</strong>.</p><ul><li><strong>Where is it?</strong> It lives right here in the monorepo inside packages/ui-kit.</li><li><strong>What is it?</strong> It is a <strong>Protocol</strong>. Think of it as a set of rules (JSON definitions) that tell apps <em>what</em> to display, without worrying about <em>how</em> it looks.</li></ul><p><strong>3. Fuselage UI Kit (</strong><strong>packages/fuselage-ui-kit)</strong>: This is also a <strong>Local Package</strong>.</p><ul><li><strong>Where is it?</strong> It lives here in packages/fuselage-ui-kit.</li><li><strong>What is it?</strong> It is the <strong>Bridge</strong>. It takes the rules from <strong>UI Kit</strong> and builds them using the components from <strong>Fuselage</strong>.</li></ul><h4>Let&#39;s focus on the main folder apps/meteor/ Where Most Code Lives</h4><p>This is the <strong>heart</strong> ❤️ of Rocket.Chat<br>Inside it:</p><ul><li>client/ : React frontend (views, components, hooks…)</li><li>server/ : Backend logic (methods, services…)</li><li>lib/ : Shared code used by both</li></ul><p>If you are debugging a UI issue, start your search here: <br>- apps/meteor/client/</p><p>If it is a backend logic issue, start your search here:<br>- apps/meteor/server/</p><h4>What helps me explore quickly:</h4><ol><li>Use global search aggressively.</li><li>Trace the flow step by step.</li><li>Follow the issue -&gt; related component -&gt; related API -&gt; server logic.</li><li>If stuck, ask in community channels.</li></ol><p>You don’t need to understand everything.<br>Even <strong>core maintainers</strong> don’t remember every part of the codebase. It is simply too large and practically impossible.<br>What really matters is understanding the <strong>flow</strong>, the <strong>structure</strong>, and the <strong>architecture</strong>.</p><p>Once you understand how things connect, navigating the monorepo becomes much easier.</p><h4>Example: Tracking a Real Issue</h4><blockquote><em>Let me give you a real example of how I navigate this monorepo.<br>I noticed the user avatar looked broken on the directory page.</em></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*U8Jal4magnHZ-1mqUzwn4w.jpeg" /></figure><blockquote><strong><em>Step 1:</em></strong><em> Search for “</em><strong><em>directory</em></strong><em>” in the codebase. You will land in<br></em>apps/meteor/client/views/directory/ ….. that is the UI.</blockquote><blockquote><strong><em>Step 2:</em></strong><em> Inside that folder, you find<br></em><strong>UsersTable</strong> the component rendering user rows with avatars.</blockquote><blockquote><strong><em>Step 3:</em></strong><em> You look at how the avatar is styled. Maybe it is a CSS issue? Maybe a prop is wrong? You read the component and compare with similar components elsewhere.</em></blockquote><blockquote><strong><em>Step 4:</em></strong><em> If the data looks wrong, you trace the API call the component makes. usually a hook like </em><strong>useQuery</strong> or <strong>useEndpoint</strong> That tells you which server endpoint to look at.</blockquote><blockquote><em>In my case, this issue was on the frontend side, so I didn’t need to go to the server side.</em></blockquote><h3><strong>What Really Matters</strong></h3><p>You don’t need to know the whole codebase. You just follow the thread from where the problem is visible to where it is caused.<br>And ask if you&#39;re stuck.</p><h4>Enjoy the open source contribution :)</h4><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=86f5f72baf9a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Understanding Solana Account Model: Everything is an Account]]></title>
            <link>https://khizarshah01.medium.com/understanding-solana-account-model-everything-is-an-account-3f86b7ccbd3d?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/3f86b7ccbd3d</guid>
            <category><![CDATA[bloackchain]]></category>
            <category><![CDATA[solana-blockchain]]></category>
            <category><![CDATA[solanas]]></category>
            <category><![CDATA[solana-network]]></category>
            <category><![CDATA[ethereum]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Thu, 19 Jun 2025 19:47:23 GMT</pubDate>
            <atom:updated>2025-06-19T19:47:23.830Z</atom:updated>
            <content:encoded><![CDATA[<p><em>A beginner’s guide to the fundamental building blocks of the Solana blockchain</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pQcKN-RmUAr7I8BSRaSO9Q.jpeg" /><figcaption>Solana Accounts</figcaption></figure><h3>Introduction</h3><p>Hey there! 👋<br>These are my notes from learning Solana development. If you’re looking for deep technical details, I recommend checking out the official Solana docs here: <a href="https://solana.com/docs/core/accounts">Solana Docs — Accounts</a></p><p>But if you’re starting and want a simple explanation, you’re in the right place!</p><p>Let’s dive in,<br>Unlike many other blockchains where smart contracts store both the code and the data together, Solana works a bit differently. It uses a unique <strong>account-based architecture</strong> where <strong>code and data are kept separate</strong>. This can be a bit confusing at first, but once it clicks, it actually makes a lot of sense.</p><p>Here’s what I’ve learned so far — explained in the simplest way possible.</p><h4>The Three Types of Accounts</h4><p>The first question that might come to your mind is: <strong>“What is an account?”</strong></p><p>On Solana, <strong>all data is stored in something called “accounts.”</strong> You can imagine Solana like a big public database, and inside that database, there’s one giant table called <strong>“Accounts.”</strong> Each row in this table represents one account.</p><p>Now, every account holds some specific information, and based on that information, we can figure out what type of account it is.<br>Here is a flow chat that simplifies the account understanding:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/879/1*GorFN_LyzvqspG4mvVg1dQ.png" /><figcaption>Flow chart of accounts</figcaption></figure><ol><li><strong>System Account/ user Accounts</strong> (Also called <em>System-Owned Accounts</em>).<br><strong>Purpose:</strong> These are your regular wallet addresses. They mainly hold <em>SOL tokens </em>(technically called <em>lamports</em>).<br><strong>Owner:</strong> Controlled by the <strong>System Program</strong> (Solana’s built-in program for basic accounts).<br><strong>Used for: </strong>Sending/receiving SOL, Paying transaction fees, Holding your wallet balance<br>&gt; <em>This is the type of account your wallet (like Phantom or Solflare) uses.</em></li><li><strong>Program Accounts</strong> (Also called Executable Accounts)<br><strong>Purpose:</strong> These accounts store the <strong>smart contract code</strong> (compiled into bytecode) that runs on Solana.<br><strong>Owner:</strong> Typically owned by the <strong>BPF Loader</strong> program, which is responsible for loading executable code onto the blockchain.<br><strong>Used for:</strong> Running smart contract logic, responding to transactions sent by users or programs.<br>&gt; <em>These accounts are marked as </em><strong><em>executable</em></strong><em>, and once deployed, the code inside them can’t be changed — unless the program was marked as </em><strong><em>upgradeable</em></strong><em> during deployment.</em></li><li><strong>Data Accounts</strong> (Also called Program-Owned Accounts)<br><strong>Purpose:</strong> These accounts store <strong>data and state</strong> for Solana programs (like smart contract variables).<br> <strong>Owner:</strong> Owned by the specific <strong>program</strong> that created them.<br> <strong>Used for:</strong> Storing things like token balances, user data, game state, or anything a program needs to remember.<br>&gt; <em>In Solana, programs themselves don’t store data — they create and manage separate accounts for that purpose.</em></li></ol><p>Now that we understand the different types of accounts, the next big question is<strong>, who controls these accounts?</strong> That’s where the concept of <em>ownership</em> comes in.</p><h3>The Ownership Model</h3><p>Every account on Solana has an <strong>owner</strong>, and this owner is always a <strong>program</strong> (not a person or wallet!).</p><p>But why does that matter?<br> Because the <strong>owner program controls</strong> what can be done with that account, just like how a mobile app controls its own data.</p><h4>Here’s the rule:</h4><ul><li>Only the <strong>owner program</strong> can change the data inside the account</li><li>Only the <strong>owner program</strong> can move SOL (lamports) out of the account</li></ul><p>This adds a strong layer of <strong>security</strong>, so random programs can’t touch your data or steal funds.</p><h4>Quick Note: What is a Token Account?</h4><p>Before we go further, in Solana, if you want to hold a token like USDC, you don’t just use your wallet address.<br> Instead, a <strong>special account called a token account</strong> is created, and it’s managed by a built-in program called the <strong>Token Program</strong>.</p><p>Your wallet <em>owns</em> the token account, but <strong>the Token Program owns and controls</strong> how that account behaves.</p><h4><strong>A Simple Example:</strong></h4><p>Let’s say you hold 10 USDC.</p><p>That 10 USDC is stored in a <strong>token account</strong> tied to your wallet, but the rules for how it works are controlled by the <strong>Token Program</strong>, not just your wallet.</p><p>Now, imagine a random smart contract tries to change your USDC balance; it <strong>can’t</strong>.</p><p>Why?<br> Because only the <strong>Token Program</strong> (the owner of that token account) has permission to modify the data inside. The Solana runtime will <strong>block any unauthorised changes</strong>.</p><h3>Rent: The Cost of Storage</h3><p>Every account on Solana must pay rent based on its size. There are two ways to handle this:</p><ul><li><strong>Rent-exempt</strong> (recommended): Pay upfront to keep the account alive forever</li><li><strong>Pay-per-epoch</strong> (rarely used): Pay rent every ~2–3 days or risk deletion</li></ul><p><strong>Formula</strong>: Larger accounts = higher rent costs</p><h3>Why This Matters</h3><p>This account model enables:</p><ul><li><strong>Separation of concerns</strong>: Code (programs) separate from data (accounts)</li><li><strong>Security</strong>: Clear ownership boundaries</li><li><strong>Efficiency</strong>: Programs can operate on multiple accounts in parallel</li><li><strong>Scalability</strong>: Data is distributed across many accounts instead of being stored in monolithic contracts</li></ul><h3>Key Takeaways</h3><ol><li>Everything in Solana is an account with an address</li><li>Accounts store raw bytes — programs interpret the structure</li><li>Three main types: user accounts, program accounts, and data accounts</li><li>Ownership determines who can modify account data</li><li>Rent prevents blockchain bloat by making storage costly</li></ol><h3>What’s Next?</h3><p>Now that I understand the account model, I’m ready to explore:</p><ul><li>Program Derived Addresses (PDAs)</li><li>Cross-Program Invocations (CPIs)</li><li>Building my first Solana program</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3f86b7ccbd3d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How DO Transactions Work on a Blockchain? (Explained for Beginners)]]></title>
            <link>https://khizarshah01.medium.com/how-do-transactions-work-on-a-blockchain-explained-for-beginners-e7407742df6c?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/e7407742df6c</guid>
            <category><![CDATA[bitcoin-transaction]]></category>
            <category><![CDATA[blockchain]]></category>
            <category><![CDATA[decentralization]]></category>
            <category><![CDATA[cryptography]]></category>
            <category><![CDATA[bitcoin]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Thu, 10 Apr 2025 05:40:17 GMT</pubDate>
            <atom:updated>2025-04-10T05:40:17.988Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Bitcoin" src="https://cdn-images-1.medium.com/max/1024/1*JaWcNyVNdXnj4D8NP29VBQ.jpeg" /></figure><h3>Introduction</h3><p>In this blog, we will explore <strong>how transactions work on a blockchain</strong> using <strong>Bitcoin</strong> as our example. Before diving into the transaction flow, we will first break down some key concepts that are essential to understand how everything works behind the scenes. These include:</p><ul><li><strong>Private Key</strong></li><li><strong>Public Key</strong></li><li><strong>Wallet</strong></li><li><strong>Miner</strong></li><li><strong>Block</strong></li></ul><p>By the end of this post, you will have a clear and simple understanding of what actually happens when someone sends or receives cryptocurrency on a blockchain.</p><h4>How Do Transactions Happen in a Bank?</h4><p>Before we get into blockchain, let’s first look at how traditional bank transactions work.</p><p>If you want to send money to a friend, you usually use an app like <strong>Google Pay</strong>, <strong>PayPal</strong>, or any other banking app. These apps are connected to your bank account. When you request to send money, the app simply:</p><ol><li><strong>Verifies your identity</strong></li><li><strong>Checks your account balance</strong></li><li><strong>Deducts the money from your account</strong></li><li><strong>Sends it to your friend’s bank account</strong></li></ol><p>It looks simple on the surface — and yes, there’s a lot happening behind the scenes — but for now, this gives us a basic idea of how centralized systems like banks handle transactions.</p><h4>How Do Transactions Happen without bank?</h4><p>In blockchain, there’s <strong>no central authority</strong> like a bank. So, how are transactions verified and processed?</p><p>That’s where <strong>miners</strong> come in. (Don’t worry — we’ll explain miners in detail later.) For now, just understand the basic flow:</p><p>Every user has a <strong>wallet</strong>, which works kind of like a PayPal or Google Pay app. When you send a transaction, it doesn’t go to a bank — it goes to the <strong>network</strong>, where <strong>miners verify your request</strong>. They check if:</p><ul><li>The transaction is really coming from you (using your <strong>digital signature</strong>)</li><li>You have enough balance in your wallet</li></ul><p>If everything looks good, the miner <strong>adds your transaction to a block</strong>, and that block becomes part of the blockchain.</p><p>Still confused? That’s okay! Before we go deeper, let’s break down the key concepts like <strong>wallets, public/private keys, miners</strong>, and <strong>blocks</strong>. Once you understand these, the entire process will make perfect sense.</p><h4>What is a Wallet?</h4><p>A <strong>wallet</strong> in the world of blockchain is just like an app — similar to PayPal or Google Pay — but with one big difference: it’s <strong>decentralized</strong>.</p><p>Let’s understand it in simple terms.</p><p>A wallet is an application that gives you a <strong>user-friendly interface</strong> to send, receive, and store cryptocurrency. It hides all the technical stuff behind the scenes so that you don’t have to worry about how it all works. You just open the app, check your balance, and send or receive crypto — easy!</p><p>But here’s the twist: the wallet itself <strong>doesn’t actually send the money</strong>. It simply <strong>creates a transaction request</strong>. That request is then sent to the <strong>blockchain network</strong>, where <strong>miners</strong> verify and process it.</p><h4>How Does a Wallet Work?</h4><p>Every wallet uses two important things:</p><h4>🔑 Public Key</h4><ul><li>Think of this like your <strong>email address</strong>.</li><li>You can <strong>share it with anyone</strong> — your friends can use it to <strong>send you crypto</strong>.</li><li>It’s also used to <strong>check your balance</strong> and see your past transactions.</li><li>Anyone can see what’s happening with a public key, but don’t worry — <strong>it doesn’t show your name, phone number, or identity</strong>. Just the key and the balance.</li></ul><h4>🔐 Private Key</h4><ul><li>This is like your <strong>password or digital signature</strong>.</li><li>You use it to <strong>sign transactions</strong> when you want to send crypto.</li><li>It proves that the transaction is really coming from <strong>you</strong>.</li><li>The private key must always be <strong>kept secret</strong>. If someone else gets it, they can control your wallet.</li></ul><p>In Short</p><ul><li>A <strong>wallet shows your balance</strong>, helps you <strong>create transactions</strong>, and keeps your <strong>keys</strong> safe.</li><li>It <strong>uses your public key</strong> to receive money and check balances.</li><li>It <strong>uses your private key</strong> to send money securely.</li><li>Once you create and sign a transaction, <strong>miners verify</strong> it and <strong>add it to the blockchain</strong>.</li></ul><h4>What are Public and Private Keys?</h4><p>Blockchain wallets don’t use usernames and passwords like traditional systems. Instead, they rely on <strong>cryptographic keys</strong>: a <strong>public key</strong> and a <strong>private key</strong>. These two keys work together to secure your crypto and allow you to send and receive transactions.</p><blockquote><strong>Public Key — Like Your Address</strong></blockquote><p>As we mentioned earlier, a <strong>public key</strong> is like your <strong>email ID or house address</strong>. You can <strong>safely share it</strong> with others so they can send you cryptocurrency.</p><p>But here’s where it’s different from normal systems:</p><ul><li>Blockchains like Bitcoin are <strong>public</strong>.</li><li>Anyone with your public key can <strong>view your wallet’s balance and transaction history</strong>.</li><li>However, your <strong>real identity remains hidden</strong> — there’s no name or mobile number attached.</li></ul><p>✅ <strong>Try it yourself</strong>:<br> Go to <a href="https://btcscan.org/">btcscan.org</a>, paste any Bitcoin public address, and you’ll see the complete transaction history and balance of that wallet. This is possible because <strong>blockchains are transparent</strong> by design.</p><blockquote><strong>Private Key — Like Your Password</strong></blockquote><p>Your <strong>private key</strong> is the most sensitive part of your wallet. Think of it like a <strong>super password</strong> or your <strong>digital signature</strong>.</p><p>Here’s what it does:</p><ul><li>It <strong>proves you are the real owner</strong> of the wallet.</li><li>It’s used to <strong>sign transactions</strong>, telling the blockchain, “Yes, I approve this.”</li><li>Without the private key, you <strong>can’t send crypto</strong> from your wallet.</li></ul><p>⚠️ <strong>Important</strong>:<br> Never share your private key.<br> Anyone who has access to it can <strong>steal all your crypto</strong>. Even the wallet app or website you use doesn’t (and shouldn’t) store your private key in plain text.</p><h3>How They Work Together</h3><p>When you send crypto:</p><ol><li>Your <strong>wallet uses your private key</strong> to sign the transaction.</li><li>The transaction is broadcast to the network.</li><li><strong>Miners or validators</strong> verify the signature using your <strong>public key</strong>.</li><li>If it’s valid, the transaction is approved and added to the blockchain.</li></ol><h3>🤔 How Do Miners Verify a Transaction If They Don’t Know the Private Key?</h3><p>You might be wondering:<br> <strong>If the private key is secret, how do miners verify that a transaction is real?</strong></p><p>Here’s the trick:</p><p>Miners don’t need to <strong>see or know</strong> your private key. Instead, your <strong>wallet uses your private key to create a digital signature</strong>.</p><p>This signature is like a <strong>mathematical proof</strong> that only your private key could have created — but it <strong>doesn’t reveal the key itself</strong>.</p><h3>🔐 What Actually Gets Sent to the Network During a Blockchain Transaction?</h3><p>When you create a transaction using your wallet, here’s what is actually sent to the blockchain network:</p><ol><li><strong>Transaction Data</strong> (the “message”):</li></ol><ul><li>Sender’s public key (your wallet address)</li><li>Receiver’s public key (friend’s wallet address)</li><li>Amount of cryptocurrency to send</li><li>Timestamp and other metadata</li></ul><p><strong>2. Digital Signature</strong> (proof of authenticity):</p><ul><li>This is generated by signing the transaction data using your <strong>private key</strong></li><li>It proves you’re the owner of the sender address, without revealing the private key</li></ul><h3>🛠️ How Miners Verify a Transaction</h3><p>Once the transaction is broadcast, <strong>miners (or validators)</strong> perform this process:</p><h4>Step 1: Receive the transaction</h4><p>The miner sees:</p><ul><li>The <strong>public key</strong> of the sender</li><li>The <strong>receiver’s address</strong></li><li>The <strong>amount</strong></li><li>The <strong>digital signature</strong></li></ul><h4>Step 2: Recalculate the signature</h4><p>Miners don’t need the private key to verify the signature. Instead, they:</p><ul><li>Use the sender’s <strong>public key</strong></li><li>Use the <strong>transaction message</strong></li><li>Use a cryptographic algorithm like <strong>ECDSA</strong> (Elliptic Curve Digital Signature Algorithm, in Bitcoin)</li></ul><h4>Step 3: Verify the signature</h4><ul><li>If the digital signature is valid for the message and public key → ✅ Transaction is real and authorized</li><li>If the signature is invalid → ❌ Miner rejects it</li></ul><h3>📦 What is a Block in Blockchain?</h3><p>A <strong>block</strong> is like a <strong>page in a digital ledger</strong>. It contains a list of valid transactions that were recently verified by the network.</p><h4>What does a block contain?</h4><p>Each block typically includes:</p><ol><li><strong>Block Header</strong> (meta information):</li></ol><ul><li>Previous Hash: The unique hash of the previous block (links blocks together)</li><li>Merkle Root: A single hash representing all transactions in the block</li><li>Timestamp: When the block was created</li><li>Nonce: A special number used during mining</li><li>Difficulty Target: Sets how hard it is to mine this block</li></ul><p>2. <strong>Transactions</strong>:</p><ul><li>All the transactions confirmed and added by the miner</li><li>Each transaction includes sender, receiver, amount, and digital signature</li></ul><h3>🔗 How blocks are connected?</h3><p>Each block <strong>references the hash of the previous block</strong>, forming a <strong>chain of blocks</strong> = <strong>Blockchain</strong><br> This ensures <strong>tamper-proof history</strong> — if one block is changed, the entire chain breaks.</p><h3>🔑 What is Hashing?</h3><p><strong>Hashing</strong> is the process of converting any input (like a transaction or block) into a <strong>fixed-size alphanumeric string</strong> using a mathematical function.</p><p>In Bitcoin, this function is called <strong>SHA-256</strong> (Secure Hash Algorithm 256-bit).</p><h4><strong>Key features of hashing:</strong></h4><ul><li>Same input → always gives same output</li><li>Tiny change in input → drastically changes the hash</li><li>One-way: you can’t reverse the hash back into the original input</li><li>Fast to compute, but impossible to guess</li></ul><pre>Input: Hello  <br>SHA-256: 185f8db32271fe25f561a6fc938b2e264306ec304eda518007d1764826381969</pre><h3>🎲 What is a Nonce?</h3><p>A <strong>nonce</strong> (short for <em>“number used once”</em>) is a number that miners keep changing to find a valid hash for a block.</p><p>Remember how we talked about <strong>hashing</strong> using SHA-256? Well, when miners create a block, they combine:</p><ul><li>🧾 Block data (transactions)</li><li>⏱️ Timestamp</li><li>🔗 Previous block hash</li><li>🔢 A number — <strong>Nonce</strong></li></ul><p>They then <strong>hash</strong> this entire block. But the hash must meet a special condition — for example:</p><blockquote><em>“The hash must start with four zeros (0000…)”</em></blockquote><h3>Let’s see an example:</h3><p>Suppose this is our block:</p><pre>Transactions: Alice → Bob = 0.5 BTC<br>Previous Hash: 0000abc123...<br>Timestamp: 2025-04-10<br>Nonce: ?</pre><p>Now the miner starts guessing different nonces:</p><ol><li><strong>Nonce = 1</strong><br> → Hash = e54fa9b02f394cc3ff91832ab9a1c3d2...<br> ❌ Doesn&#39;t start with 0000 — Not valid</li><li><strong>Nonce = 2</strong><br> → Hash = a3b1c9820a2dd33fc09d91d9a77fe812...<br> ❌ Still not valid</li><li><strong>Nonce = 3</strong><br> → Hash = 0000bc1298b9d91c38aa2de0f3b8f22d...<br> ✅ Success! Hash starts with 0000</li></ol><h3>This process is called Proof of Work</h3><p>Miners try <strong>millions or even billions of nonces</strong> until someone finds the right one. It takes <strong>real computational effort</strong>, which is why mining rewards cost energy.</p><h3>Why is this important?</h3><ul><li>It keeps the blockchain secure</li><li>It prevents spamming or fake blocks</li><li>It ensures only honest work adds new blocks</li></ul><h3>✅ How the Transaction Becomes Final</h3><p>Once a transaction is verified by a miner (using your <strong>digital signature</strong> and the <strong>private key</strong>), it doesn’t become “final” immediately.</p><p>Here’s what happens:</p><ol><li>✅ <strong>Transaction gets added to a block</strong> by a miner.</li><li>⛓️ That block is added to the blockchain (linked to the previous block).</li><li>⏳ <strong>Other miners start adding more blocks on top</strong> of it.</li><li>✅ Each new block added = 1 <strong>confirmation</strong>.</li></ol><blockquote><em>💡 The more confirmations a transaction has, the more secure it becomes.<br> For example:</em></blockquote><ul><li>1 confirmation: Block added</li><li>6 confirmations: Considered final in Bitcoin (very hard to reverse)</li></ul><p>Once your transaction is <strong>deep enough in the chain</strong>, it becomes <strong>practically irreversible</strong>, because changing it would require re-mining all the blocks after it — which is computationally impossible.</p><h3>✅ Conclusion: Why It’s Secure and Trustless</h3><p>Let’s quickly recap how everything fits together:</p><ul><li>🔐 <strong>Wallets</strong> give you a private key (to sign) and a public key (to receive).</li><li>✅ Transactions are digitally signed, so only you can authorize sending crypto.</li><li>⛏️ <strong>Miners</strong> verify your signature and include the transaction in a block.</li><li>🔄 <strong>Hashing &amp; Nonce</strong> make sure blocks are valid and secure.</li><li>⛓️ Once added, each block is connected to the previous, forming a chain.</li><li>✔️ The more confirmations a transaction gets, the more permanent it becomes.</li></ul><blockquote><em>🔒 No central authority. No one to trust. Just </em><strong><em>math, code, and proof</em></strong><em>.</em></blockquote><p>That’s the power of blockchain — <strong>secure, transparent, and trustless</strong> by design.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e7407742df6c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kadane’s Algorithm: Maximum Subarray Sum Made Easy]]></title>
            <link>https://khizarshah01.medium.com/kadanes-algorithm-maximum-subarray-sum-made-easy-35980a30c16e?source=rss-58044a6b4850------2</link>
            <guid isPermaLink="false">https://medium.com/p/35980a30c16e</guid>
            <category><![CDATA[kadanes-algorithm]]></category>
            <category><![CDATA[java]]></category>
            <category><![CDATA[data-structure-algorithm]]></category>
            <dc:creator><![CDATA[Khizar Shah]]></dc:creator>
            <pubDate>Wed, 19 Feb 2025 16:23:45 GMT</pubDate>
            <atom:updated>2025-02-19T16:23:45.431Z</atom:updated>
            <content:encoded><![CDATA[<p>A common problem in coding interviews and competitive programming is finding the <strong>maximum sum of a contiguous subarray</strong>. Instead of checking all possible subarrays (which take <strong>O(n²)</strong> time), we can solve this efficiently using <strong>Kadane’s Algorithm</strong> in <strong>O(n) time</strong>.</p><h3>Understanding the Problem</h3><p>Given an array of integers, find the contiguous subarray with the <strong>largest sum</strong>.</p><blockquote><em>Example:</em></blockquote><p>Input: [1, -3, 4, -1, 2, 1, -5, 4]</p><p>Output: 6</p><p>Explanation: The subarray [4, -1, 2, 1] gives the maximum sum: 4+(−1)+2+1= 6</p><h3>Kadane’s Algorithm (O(n) Approach)</h3><p>Kadane’s Algorithm keeps track of:<br>✔ currentSum → Maximum sum ending at the current position.<br>✔ maxSum → Global maximum sum found so far.</p><blockquote>Algorithm Steps:</blockquote><p>Step 1: Initialize maxSum = arr[0] and currentSum = arr[0].</p><p>Step 2: Loop through the array from the index 1 to n-1:</p><ul><li>Update currentSum = max(arr[i], currentSum + arr[i]).</li><li>Update maxSum = max(maxSum, currentSum).</li></ul><p>Step 4: Return maxSum.</p><h3>1. Extend the current subarray</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/932/1*aSuU6-k6kDZ1emSOJPwclw.png" /></figure><h3>2. Start fresh with a new subarray</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/937/1*Nty38XBVGbJIINJ91AwC1g.png" /></figure><h3>Kadane’s Algorithm Implementation in Java</h3><pre>public class KadanesAlgorithm {<br>    public static int maxSubarraySum(int[] nums) {<br>        int maxSum = nums[0];<br>        int currentSum = nums[0];<br><br>        for (int i = 1; i &lt; nums.length; i++) {<br>            currentSum = Math.max(nums[i], currentSum + nums[i]);<br>            maxSum = Math.max(maxSum, currentSum);<br>        }<br><br>        return maxSum;<br>    }<br><br>    public static void main(String[] args) {<br>        int[] arr = {1, -3, 4, -1, 2, 1, -5, 4};<br>        System.out.println(&quot;Maximum Subarray Sum: &quot; + maxSubarraySum(arr));<br>    }<br>}</pre><h3>Time &amp; Space Complexity:</h3><ul><li><strong>O(n) Time Complexity</strong> — Single pass through the array.</li><li><strong>O(1) Space Complexity</strong> — Uses only a few variables.</li></ul><h3>Recommended Video for Better Understanding</h3><p>If you prefer visual explanations, check out this short and clear video on Kadane’s Algorithm:</p><p>📌 <a href="https://youtu.be/NUWAXbSlsws?si=apGjaqvK4MvArEq4">Kadane’s Algorithm Explained</a></p><h3>Conclusion</h3><p>Kadane’s Algorithm is one of the most efficient ways to find the <strong>maximum subarray sum</strong> in <strong>O(n) time</strong>. It’s a must-know for <strong>coding interviews and competitive programming</strong>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=35980a30c16e" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>