<?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 Anuja Lamahewa on Medium]]></title>
        <description><![CDATA[Stories by Anuja Lamahewa on Medium]]></description>
        <link>https://medium.com/@anujalamahewa?source=rss-58afef076ad2------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*d4s_l5wo3i9e3aMbsLjjpw.jpeg</url>
            <title>Stories by Anuja Lamahewa on Medium</title>
            <link>https://medium.com/@anujalamahewa?source=rss-58afef076ad2------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 14:35:08 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@anujalamahewa/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[Monolithic vs. Microservices: Choosing the Right Architecture]]></title>
            <link>https://anujalamahewa.medium.com/monolithic-vs-microservices-choosing-the-right-architecture-b71266bac553?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/b71266bac553</guid>
            <category><![CDATA[microservices]]></category>
            <category><![CDATA[monolithic]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Thu, 29 Feb 2024 00:23:58 GMT</pubDate>
            <atom:updated>2024-03-14T00:33:09.338Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/765/1*UoWW5dgrfjvKOadps2VSFg.gif" /></figure><p>In the ever-evolving landscape of software development, architectural paradigms often come and go, each promising to revolutionize the way we build and deploy applications. One such paradigm that gained traction in recent years is microservices architecture. Initially, microservices seemed like the silver bullet we’ve been waiting for — small, nimble services that promised easy reasoning and scalability within the boundaries of each service.</p><p>However, as many of us have experienced firsthand, the journey from monoliths to microservices is not without its challenges. The first hurdle we encountered was the complexity of managing communication between microservices. API contracts, unreliable networks, and services crossing domain boundaries added layers of complexity that we hadn’t fully anticipated.</p><p>To address these challenges, we introduced various solutions — from implementing contracts across services to adopting technologies like GraphQL, gRPC, and asynchronous communication. We found ourselves grappling with issues such as message queues, retries with exponential backoffs, and the management of shared, versioned libraries. In the quest for scalability and flexibility, we inadvertently created what some have termed a “distributed monolith.”</p><p>Moreover, the proliferation of microservices led to a proliferation of source code repositories, each with its own slightly different patterns and practices. This made it increasingly difficult to maintain consistency across our codebase and keep up with necessary patches and updates. In response, we turned to monorepos in an attempt to centralize our codebase and streamline development processes.</p><p>Yet, despite our best efforts to avoid the pitfalls of monolithic architectures, we found ourselves grappling with a new set of challenges. The overhead of managing a distributed system, coupled with the complexity of coordinating deployments and versioning, often felt like a drag on productivity. What once felt like a promising solution to the woes of monoliths now seemed to introduce its own set of complexities and inefficiencies.</p><p>As we reflect on our journey from monoliths to microservices, it’s clear that there are no easy answers or one-size-fits-all solutions. While microservices offer undeniable benefits in terms of scalability and resilience, they also come with inherent complexities that cannot be ignored. Moving forward, it’s essential to strike a balance between the benefits of microservices and the practical realities of managing a distributed system.</p><p>In conclusion, the transition from monoliths to microservices is not a panacea, but rather a journey fraught with challenges and trade-offs. By acknowledging the limitations of microservices and adopting a pragmatic approach to architecture and development, we can navigate the complexities of modern software engineering more effectively and ultimately deliver better outcomes for our teams and users.</p><p>Understanding Monolithic Architecture:</p><p>Imagine a colossal, all-encompassing structure where every component of an application is tightly integrated and interconnected. This is the essence of a monolithic architecture. In this paradigm, the entire application is developed, deployed, and scaled as a single unit.</p><p>Advantages of Monolithic Architecture:</p><ol><li>Simplicity: Monolithic architectures are relatively straightforward to develop and deploy since the entire application is contained within a single codebase.</li><li>Ease of Testing: Testing in a monolithic architecture is more straightforward as all components are tightly coupled, facilitating comprehensive end-to-end testing.</li><li>Performance: Inter-component communication within a monolith occurs via method calls or function invocations, which are inherently fast compared to network calls in a distributed system.</li></ol><p>Drawbacks of Monolithic Architecture:</p><ol><li>Scalability Challenges: Scaling a monolithic application involves scaling the entire system, even if only a specific component requires additional resources. This can lead to inefficiencies and increased infrastructure costs.</li><li>Limited Technology Stack Flexibility: Since all components share the same codebase and runtime environment, adopting new technologies or frameworks becomes cumbersome and risky.</li><li>Maintenance Complexity: As monolithic applications grow larger, maintaining and updating them becomes increasingly complex and time-consuming.</li></ol><p>Understanding Microservices Architecture:</p><p>Contrastingly, Microservices architecture decomposes an application into a collection of small, independently deployable services, each responsible for a specific business function. These services communicate with each other via lightweight protocols like HTTP or messaging queues.</p><p>Advantages of Microservices Architecture:</p><ol><li>Scalability: Microservices offer granular scalability, allowing individual components to be scaled independently based on demand. This leads to optimized resource utilization and cost-effectiveness.</li><li>Technology Diversity: Each microservice can be developed using the most suitable technology stack for its specific requirements, promoting flexibility and innovation.</li><li>Isolation and Fault Tolerance: Since microservices operate independently, failures in one service do not necessarily affect the entire system, enhancing fault tolerance and resilience.</li></ol><p>Drawbacks of Microservices Architecture:</p><ol><li>Complexity: The distributed nature of microservices introduces complexity in terms of service discovery, inter-service communication, and data consistency.</li><li>Operational Overhead: Managing a large number of microservices entails significant operational overhead, including deployment automation, monitoring, and debugging.</li><li>Consistency Challenges: Ensuring data consistency and transactional integrity across multiple services can be challenging, especially in distributed transactions.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/580/1*hUxisvJh4P8kpRJ1_SxVTg.gif" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b71266bac553" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Simple Overview of Shipping Code]]></title>
            <link>https://anujalamahewa.medium.com/simple-overview-of-shipping-code-87ff00e77866?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/87ff00e77866</guid>
            <category><![CDATA[release-management]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Wed, 28 Feb 2024 14:46:49 GMT</pubDate>
            <atom:updated>2024-03-14T00:35:21.734Z</atom:updated>
            <content:encoded><![CDATA[<p>The diagram below illustrates the typical workflow of shipping the product to the customer:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/756/1*P4xq7F0VRP50Y-_8fRpXIw.gif" /></figure><p><strong>Step 1:</strong> User stories are essential for capturing end-user requirements and defining the scope of development work. The product owner collaborates closely with stakeholders to ensure these stories accurately reflect user needs and business objectives. Clear, concise user stories facilitate effective communication between the development team and stakeholders throughout the development process.</p><p><strong>Step 2:</strong> The two-week sprint cycle provides a structured timeframe for development efforts, promoting focus and productivity. Within each sprint, the development team commits to completing a set of user stories, fostering accountability and progress tracking. Regular sprint planning meetings allow the team to prioritize tasks and allocate resources effectively.</p><p><strong>Step 3:</strong> Git is a widely used version control system that enables collaborative development and code management. Developers leverage Git’s branching and merging capabilities to work on feature branches, facilitating parallel development efforts while maintaining code integrity. Committing source code to the repository ensures that changes are tracked, documented, and accessible to team members.</p><p><strong>Step 4:</strong> Jenkins is an automation server used for continuous integration and continuous delivery (CI/CD). It automates the build and deployment process, streamlining code integration and validation. By triggering builds automatically whenever changes are pushed to the repository, Jenkins enables rapid feedback on code quality and ensures that only stable, tested code progresses through the pipeline.</p><p><strong>Step 5:</strong> Artifactory serves as a binary repository manager, storing artifacts generated during the build process. Storing builds in Artifactory ensures artifact traceability and reproducibility, facilitating efficient artifact management and deployment. Deploying builds to the development environment allows developers to validate changes in a controlled setting before moving to subsequent stages of testing.</p><p><strong>Step 6:</strong> Deploying features to separate QA environments enables isolated testing of individual features, preventing interference between different development efforts. QA1 and QA2 environments provide dedicated testing environments for quality assurance teams to conduct comprehensive testing, including functional testing, regression testing, and performance testing.</p><p><strong>Step 7:</strong> The QA team plays a crucial role in ensuring the quality and reliability of software releases. By conducting rigorous testing in dedicated QA environments, QA engineers identify and address defects, usability issues, and performance bottlenecks before software is deployed to production. Thorough testing helps mitigate the risk of introducing bugs or regressions in production environments.</p><p><strong>Step 8: </strong>Deploying features involves releasing new functionality or changes to a software system in a controlled manner, often using techniques like feature branching, feature flags, and feature rollouts to isolate and gradually introduce updates. Canary releases and blue-green deployments enable developers to test changes in production environments while minimizing risk and downtime. Automated deployment pipelines streamline the process, ensuring rapid and consistent deployments from development through testing to production. Monitoring and rollback strategies are crucial for detecting issues and reverting changes if necessary, helping organizations deliver new features confidently while maintaining the stability and reliability of their software systems.</p><p><strong>Step 9: </strong>User Acceptance Testing (UAT) serves as the final validation phase before software is released to production. UAT environments replicate the production environment closely, allowing stakeholders to evaluate the software’s functionality and usability in a real-world setting. Successful completion of UAT indicates that the software meets user expectations and is ready for production deployment.</p><p><strong>Step 10:</strong> Once UAT testing is complete and the software is deemed ready for release, it transitions to the production environment. Release candidates undergo final deployment procedures, including configuration management and rollout planning, to ensure a smooth transition to production. Coordination between development, operations, and release management teams is critical to minimize downtime and mitigate deployment risks.</p><p><strong>Step 11:</strong> The Site Reliability Engineering (SRE) team is responsible for ensuring the reliability, availability, and performance of production systems. SREs monitor production environments proactively, using tools and techniques to detect and mitigate incidents before they impact end users. By implementing best practices in infrastructure management and automation, SREs help maintain the stability and scalability of production systems.</p><h3>Ways to Test the System</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*_oQwqMwGuJE-r90ZMwakZg.gif" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=87ff00e77866" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[System Design Core Concepts]]></title>
            <link>https://anujalamahewa.medium.com/system-design-core-concepts-860a30ca41c2?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/860a30ca41c2</guid>
            <category><![CDATA[system-design-concepts]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Tue, 27 Feb 2024 06:02:14 GMT</pubDate>
            <atom:updated>2024-02-27T06:02:14.193Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*3lHvaHgyrbBQ2xhk.gif" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=860a30ca41c2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Functional & Non-Functional test]]></title>
            <link>https://anujalamahewa.medium.com/functional-non-functional-test-60314136a181?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/60314136a181</guid>
            <category><![CDATA[functional-testing]]></category>
            <category><![CDATA[testing]]></category>
            <category><![CDATA[non-functional-testing]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Tue, 27 Feb 2024 05:56:48 GMT</pubDate>
            <atom:updated>2024-02-27T05:56:48.129Z</atom:updated>
            <content:encoded><![CDATA[<h3>Functional &amp; Non-Functional Testing</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/792/0*WfuEVJUpIFP_2WAX.gif" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=60314136a181" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Project Management Documentation Templates]]></title>
            <link>https://anujalamahewa.medium.com/project-management-documentation-templates-f6a57862d013?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/f6a57862d013</guid>
            <category><![CDATA[project-management]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Tue, 27 Feb 2024 05:51:17 GMT</pubDate>
            <atom:updated>2024-02-27T05:51:17.501Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*ecMAGREhm93ECOe5" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f6a57862d013" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Essential Git Commands]]></title>
            <link>https://anujalamahewa.medium.com/essential-git-commands-95dc1ab63c48?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/95dc1ab63c48</guid>
            <category><![CDATA[git]]></category>
            <category><![CDATA[version-control]]></category>
            <category><![CDATA[github]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Tue, 20 Feb 2024 22:57:12 GMT</pubDate>
            <atom:updated>2024-03-04T23:55:41.934Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/0*89JgRQHcCltOyr9J.gif" /></figure><p>- 𝗜𝗻𝗶𝘁𝗶𝗮𝗹𝗶𝘇𝗲 𝗥𝗲𝗽𝗼𝘀𝗶𝘁𝗼𝗿𝘆: 𝚐𝚒𝚝 𝚒𝚗𝚒𝚝<br>- 𝗖𝗹𝗼𝗻𝗲 𝗥𝗲𝗽𝗼𝘀𝗶𝘁𝗼𝗿𝘆: 𝚐𝚒𝚝 𝚌𝚕𝚘𝚗𝚎 [𝚄𝚁𝙻]</p><p>Working with Remotes</p><p>- 𝗟𝗶𝘀𝘁 𝗥𝗲𝗺𝗼𝘁𝗲𝘀: 𝚐𝚒𝚝 𝚛𝚎𝚖𝚘𝚝𝚎 -𝚟<br>- 𝗔𝗱𝗱 𝗥𝗲𝗺𝗼𝘁𝗲: 𝚐𝚒𝚝 𝚛𝚎𝚖𝚘𝚝𝚎 𝚊𝚍𝚍 [𝚗𝚊𝚖𝚎] [𝚄𝚁𝙻]<br>- 𝗥𝗲𝗺𝗼𝘃𝗲 𝗥𝗲𝗺𝗼𝘁𝗲: 𝚐𝚒𝚝 𝚛𝚎𝚖𝚘𝚝𝚎 𝚛𝚖 [𝚗𝚊𝚖𝚎]<br>- 𝗙𝗲𝘁𝗰𝗵 𝗥𝗲𝗺𝗼𝘁𝗲 𝗖𝗵𝗮𝗻𝗴𝗲𝘀: 𝚐𝚒𝚝 𝚏𝚎𝚝𝚌𝚑 [𝚛𝚎𝚖𝚘𝚝𝚎]<br>- 𝗣𝘂𝗹𝗹 𝗖𝗵𝗮𝗻𝗴𝗲𝘀: 𝚐𝚒𝚝 𝚙𝚞𝚕𝚕 [𝚛𝚎𝚖𝚘𝚝𝚎] [𝚋𝚛𝚊𝚗𝚌𝚑]️<br>- 𝗣𝘂𝘀𝗵 𝗖𝗵𝗮𝗻𝗴𝗲𝘀: 𝚐𝚒𝚝 𝚙𝚞𝚜𝚑 [𝚛𝚎𝚖𝚘𝚝𝚎] [𝚋𝚛𝚊𝚗𝚌𝚑]</p><p>Branching &amp; Merging</p><p>- 𝗟𝗶𝘀𝘁 𝗕𝗿𝗮𝗻𝗰𝗵𝗲𝘀: 𝚐𝚒𝚝 𝚋𝚛𝚊𝚗𝚌𝚑<br>- 𝗖𝗿𝗲𝗮𝘁𝗲 𝗕𝗿𝗮𝗻𝗰𝗵: 𝚐𝚒𝚝 𝚋𝚛𝚊𝚗𝚌𝚑 [𝚗𝚊𝚖𝚎]<br>- 𝗦𝘄𝗶𝘁𝗰𝗵 𝗕𝗿𝗮𝗻𝗰𝗵: 𝚐𝚒𝚝 𝚌𝚑𝚎𝚌𝚔𝚘𝚞𝚝 [𝚗𝚊𝚖𝚎]<br>- 𝗖𝗿𝗲𝗮𝘁𝗲 &amp; 𝗦𝘄𝗶𝘁𝗰𝗵 𝗕𝗿𝗮𝗻𝗰𝗵: 𝚐𝚒𝚝 𝚌𝚑𝚎𝚌𝚔𝚘𝚞𝚝 -𝚋 [𝚗𝚊𝚖𝚎]<br>- 𝗠𝗲𝗿𝗴𝗲 𝗕𝗿𝗮𝗻𝗰𝗵: 𝚐𝚒𝚝 𝚖𝚎𝚛𝚐𝚎 [𝚋𝚛𝚊𝚗𝚌𝚑]<br>- 𝗗𝗲𝗹𝗲𝘁𝗲 𝗕𝗿𝗮𝗻𝗰𝗵: 𝚐𝚒𝚝 𝚋𝚛𝚊𝚗𝚌𝚑 -𝚍 [𝚗𝚊𝚖𝚎]</p><p>Making Changes</p><p>- 𝗦𝘁𝗮𝘁𝘂𝘀: 𝚐𝚒𝚝 𝚜𝚝𝚊𝚝𝚞𝚜<br>- 𝗔𝗱𝗱 𝗖𝗵𝗮𝗻𝗴𝗲𝘀: 𝚐𝚒𝚝 𝚊𝚍𝚍 [𝚏𝚒𝚕𝚎/𝚍𝚒𝚛𝚎𝚌𝚝𝚘𝚛𝚢]<br>- 𝗖𝗼𝗺𝗺𝗶𝘁 𝗖𝗵𝗮𝗻𝗴𝗲𝘀: 𝚐𝚒𝚝 𝚌𝚘𝚖𝚖𝚒𝚝 -𝚖 “[𝚖𝚎𝚜𝚜𝚊𝚐𝚎]”<br>- 𝗔𝗺𝗲𝗻𝗱 𝗖𝗼𝗺𝗺𝗶𝘁: 𝚐𝚒𝚝 𝚌𝚘𝚖𝚖𝚒𝚝 — 𝚊𝚖𝚎𝚗𝚍<br>- 𝗥𝗲𝘀𝗲𝘁: 𝚐𝚒𝚝 𝚛𝚎𝚜𝚎𝚝 [𝚏𝚒𝚕𝚎]<br>- 𝗛𝗮𝗿𝗱 𝗥𝗲𝘀𝗲𝘁: 𝚐𝚒𝚝 𝚛𝚎𝚜𝚎𝚝 — 𝚑𝚊𝚛𝚍 [𝚌𝚘𝚖𝚖𝚒𝚝]</p><p>Reviewing History</p><p>- 𝗟𝗼𝗴: 𝚐𝚒𝚝 𝚕𝚘𝚐<br>- 𝗟𝗼𝗴 (𝗚𝗿𝗮𝗽𝗵): 𝚐𝚒𝚝 𝚕𝚘𝚐 — 𝚐𝚛𝚊𝚙𝚑<br>- 𝗦𝗵𝗼𝘄 𝗖𝗵𝗮𝗻𝗴𝗲: 𝚐𝚒𝚝 𝚜𝚑𝚘𝚠 [𝚌𝚘𝚖𝚖𝚒𝚝]<br>- 𝗗𝗶𝗳𝗳: 𝚐𝚒𝚝 𝚍𝚒𝚏𝚏 [𝚋𝚛𝚊𝚗𝚌𝚑]</p><p>Cleanup &amp; Maintenance</p><p>- 𝗦𝘁𝗮𝘀𝗵 𝗖𝗵𝗮𝗻𝗴𝗲𝘀: 𝚐𝚒𝚝 𝚜𝚝𝚊𝚜𝚑<br>- 𝗔𝗽𝗽𝗹𝘆 𝗦𝘁𝗮𝘀𝗵: 𝚐𝚒𝚝 𝚜𝚝𝚊𝚜𝚑 𝚙𝚘𝚙<br>- 𝗖𝗹𝗲𝗮𝗻 𝗨𝗻𝘁𝗿𝗮𝗰𝗸𝗲𝗱 𝗙𝗶𝗹𝗲𝘀: 𝚐𝚒𝚝 𝚌𝚕𝚎𝚊𝚗 -𝚏</p><p>Advanced &amp; Miscellaneous</p><p>- 𝗥𝗲𝗯𝗮𝘀𝗲: 𝚐𝚒𝚝 𝚛𝚎𝚋𝚊𝚜𝚎 [𝚋𝚛𝚊𝚗𝚌𝚑]<br>- 𝗖𝗵𝗲𝗿𝗿𝘆-𝗽𝗶𝗰𝗸: 𝚐𝚒𝚝 𝚌𝚑𝚎𝚛𝚛𝚢-𝚙𝚒𝚌𝚔 [𝚌𝚘𝚖𝚖𝚒𝚝]<br>- 𝗧𝗮𝗴: 𝚐𝚒𝚝 𝚝𝚊𝚐 [𝚗𝚊𝚖𝚎]<br>- 𝗦𝗲𝗮𝗿𝗰𝗵 𝗟𝗼𝗴: 𝚐𝚒𝚝 𝚕𝚘𝚐 — 𝚐𝚛𝚎𝚙=”[𝚙𝚊𝚝𝚝𝚎𝚛𝚗]”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PcYF-0e5wodnO_GNfnO26Q.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*viG73KcDcUopZ2pAWqs-XA.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kan7AhDsuKXO7EVfSRecrQ.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9Ut8EeLMrFXy7MMHdioNTA.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Df8X3FYfqprF2P-2QsgVDg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*C2-aaiIWbW8bZydn_K-XjQ.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tA9ogeXCQTFsL4GeImnz1A.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LrhGLw4L4QxKaAAcYwLE7Q.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=95dc1ab63c48" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Introduction to Vision Transformer]]></title>
            <link>https://anujalamahewa.medium.com/introduction-to-vision-transformer-1cb8265a1d1a?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/1cb8265a1d1a</guid>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Tue, 17 Oct 2023 14:24:58 GMT</pubDate>
            <atom:updated>2024-02-20T23:06:11.053Z</atom:updated>
            <content:encoded><![CDATA[<p>Hey there! Let’s dive into the fascinating world of vision transformers. These cutting-edge models have revolutionized the field of computer vision by introducing a new approach to image recognition and understanding.</p><p>In this article, we’ll explore what vision transformers are and how they have become a game-changer in the world of artificial intelligence. So, buckle up and get ready to unravel the magic behind these incredible algorithms!</p><p><strong>Humble Beginnings</strong></p><p>The Vision Transformer model was introduced in 2021 in a conference research paper titled “<a href="https://arxiv.org/pdf/2010.11929.pdf">An Image is Worth 16*16 Words: Transformers for Image Recognition at Scale</a>,” published at ICLR 2021.</p><p><strong>What is a Vision Transformer?</strong></p><p>A Vision Transformer is a type of neural network that can be used for image classification and other computer vision tasks.</p><p>One of the key changes that Vision Transformers make is the way that they <strong>represent images</strong>. In natural language processing, Transformer models typically represent text as a sequence of words. However, images cannot be represented as a sequence of words. Instead, Vision Transformers represent images as a sequence of patches.</p><p><strong>Applications</strong></p><p>Vision transformers have extensive applications in popular image recognition tasks such as object detection, image segmentation, image classification, and action recognition. Moreover, ViTs are applied in generative modeling and multi-model tasks, including visual grounding, visual question answering, and visual reasoning.</p><p><strong>How Do Vision Transformers Work?</strong></p><p>Images are represented as sequences, and class labels for the image are predicted, which allows models to learn image structure independently. Input images are treated as a sequence of patches where every patch is flattened into a single vector by concatenating the channels of all pixels in a patch and then linearly projecting it to the desired input dimension.</p><p><strong>Vision Transformer Architecture</strong></p><p>We can break into the following steps:</p><ol><li>Split an image into <strong>patches </strong><em>(A patch is a small rectangular region of an image. The size of the patch is typically 16x16 pixels).</em></li><li>Flatten the patches.</li><li>Produce lower-dimensional linear embeddings from the flattened patches.</li><li>Add positional embeddings.</li><li>Feed the sequence as an input to a standard transformer encoder.</li><li>Pretrain the model with image labels <em>(fully supervised on a huge dataset).</em></li><li>Finetune on the downstream dataset for image classification.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ZkQyuVJkf9KNU5v9.gif" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1cb8265a1d1a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Supervised, Unsupervised andSemi Supervisedlearning]]></title>
            <link>https://anujalamahewa.medium.com/supervised-unsupervised-andsemi-supervisedlearning-70095952e2f4?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/70095952e2f4</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[semi-supervised-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[supervised-learning]]></category>
            <category><![CDATA[unsupervised-learning]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Mon, 09 Oct 2023 03:12:52 GMT</pubDate>
            <atom:updated>2023-10-09T03:41:54.282Z</atom:updated>
            <content:encoded><![CDATA[<h3>Supervised, Unsupervised and Semi-Supervised Learning</h3><p>Supervised and unsupervised learning are two core components in building machine learning models.</p><p>So, what’s the difference?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MNF7qC9ru7mKLHxXcCZRJQ.png" /></figure><p><strong>Supervised Learning</strong></p><p>supervised learning, which uses labelled input and output data, while an unsupervised learning model doesn’t.</p><p>But what does that really mean?</p><p>Now, in supervised learning, the machine learning algorithm is trained on a labelled dataset. So, this means that for each example in the training dataset, the algorithm knows what the correct output is, and the algorithm uses this knowledge to try to generalize to new examples that it’s never seen before. Thus, using labelled inputs and outputs, the model can measure its accuracy and learn over time.</p><p>Supervised learning can be divided into a couple of subcategories.</p><p><strong>Classification</strong></p><p>Firstly, there is a category of classification and classification talks about whether the output is a discrete class label such as “spam” or “not spam”. Linear classifiers, support vector machines, or SPMs, decision trees, and random forests which all common examples of classification algorithms.</p><p><strong>Regression</strong></p><p>The other example is regression. The output here is a continuous value, such as price or probability. Linear regression and logistic regression are two common types of regression algorithms.</p><p><strong>Unsupervised Learning</strong></p><p>Now, unsupervised learning is where the machine learning algorithm is not really given any labels at all, and these algorithms discover hidden patterns in data without the need for human intervention. They’re unsupervised.</p><p>Unsupervised learning models are used for three main tasks, such as clustering, association, and dimensionality reduction. So, let’s take a look at each one of those, starting with clustering.</p><p><strong>Clustering</strong></p><p>Clustering is where the algorithm groups similar experiences together. So, a common application of clustering is customer segmentation, where businesses might group customers together based on similarities like, I don’t know, age or location or spending habits, something like that.</p><p><strong>Association</strong></p><p>And association is where the algorithm looks for relationships between variables in the data. Association rules are often used in market basket analysis, where businesses want to know which items are often bought together. You know, something along the lines of, “Customers who bought this item also bought “, that sort of thing.</p><p><strong>Dimensional Reduction</strong></p><p>And this is where the algorithm reduces the number of variables in the data, while still preserving as much of the information as possible. Often this technique is used in the pre-processing data stage, such as when autoencoders remove noise from visual images to improve picture quality.</p><p><strong>Differences between these two types of learning</strong></p><p>Okay, so let’s talk about the differences between these two types of learning.</p><p>In supervised learning, the algorithm learns from training datasets by iteratively making predictions on the data and then adjusting for the correct answer.</p><p>While supervised learning models tend to be more accurate than unsupervised learning models,</p><p>they do require all this up-front human intervention to label the data appropriately. For example, a supervised learning model can predict how long your commute will be on the time of day and think about the weather conditions and so forth but first, you’ll have to train it to know things like rainy weather extends the driving time.</p><p>By contrast, unsupervised learning models work on their own to discover the inherent structure of unlabelled data. These models don’t need humans to intervene. They can automatically find patterns in data and group them together.</p><p>So, for example, an unsupervised learning model can cluster images by the objects they contain things like people and animals and buildings without being told what those objects were ahead of time.</p><p>Now, an important distinction to make is that unsupervised learning models don’t make predictions. They only group data together. So if you were to use an unsupervised learning model on that same commute dataset, it would group together commutes with similar conditions like the time of day and the weather, but it wouldn’t be able to predict how long each commute would take.</p><p>Okay, so which of these two options is right for you?</p><p>In general, supervised learning is more commonly used than unsupervised learning and that’s really because it’s more accurate and efficient, that being said, unsupervised learning has its own advantages. There are two that I can think of.</p><p>Firstly, unsupervised learning can be used on data that is not labelled, which is often the case in real-world datasets.</p><p>And then secondly, unsupervised learning can be used to find hidden patterns in data that supervised learning models just wouldn’t find.</p><p>Classifying big data can be a real challenge in supervised learning, but the results are highly accurate and trustworthy in contrast, unsupervised learning can handle large volumes of data in real-time but there’s a lack of transparency into how that data is clustered and a high risk given accurate results.</p><p>But wait, it is not an “either/or” choice.</p><p>May I present to you the middle ground known as semi-supervised learning:</p><p><strong>Semi-Supervised Learning</strong></p><p>This is, well, a happy medium where you use a training data set with both labelled and unlabelled data. It’s particularly useful when it’s difficult to extract relevant features from data when you have a high volume of data. So, for example, you could use a semi-supervised learning algorithm on a data set with millions of images where only a few thousand of those images are labelled.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jxJZejDmzEigD2afg5KwuQ.png" /></figure><p>Semi-supervised learning is ideal for medical images, where a small amount of training data could lead to a significant improvement in accuracy. For example, a radiologist can look at and label some small subset of CT scans for tumours or diseases, and then the machine can more accurately predict which patients might require more medical attention without going through and labelling the entire set.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/851/1*wfGCK-IkSDC2D23XxyQDtw.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=70095952e2f4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Ultimate Guide to managing media in iPhone]]></title>
            <link>https://anujalamahewa.medium.com/ultimate-guide-to-managing-media-in-iphone-720f848d64af?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/720f848d64af</guid>
            <category><![CDATA[ios]]></category>
            <category><![CDATA[apple]]></category>
            <category><![CDATA[android]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Tue, 27 Oct 2015 04:56:52 GMT</pubDate>
            <atom:updated>2015-10-27T07:46:19.350Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>iSheep</strong></h3><p>“ A follower of the Apple cult regardless of the usefulness or real worth of the product. Believes with without question the cult propaganda which installs the almost mythological belief that what they have just bought is the fastest or most user friendly product ever, only to be re-sold the same product, with a few minor tweaks, a few months later with the same rhetoric and complete disregard for reality.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/650/1*j2qQuxpdNH03MN3zOt2nwQ.jpeg" /><figcaption>Oh come on! I had to put this picture</figcaption></figure><p>Oh let’s start. I’m getting a feeling that this will be a long post..</p><p>This is my second post regarding iPhone.</p><h3>About me:</h3><p>I was a long time hard core Android user specialized (i would say) in rooting and Android customization. The more and more deep you go into the root world the more and more you understand how the whole memory management in Android works (CPU Governors and I/O Schedulers). This is the problem, my opinion is Android needs to change the way things work and people like Samsung should not force their own rules when it comes to I/O Schedulers… You know this should be a separate article. Anyway Android user then bought the iPhone 6S. There now you’re updated about me.</p><p>Now we can move on…</p><p>WARNING: You have to buy third party software if you keep reading this article!</p><p>Okey so managing ANY sort of media is a nightmare in iOS. I understand the security bla bla where Apple doesn’t give access to it’s folders but this it’s stupid and not practical (especially if you’re a former Android user who’s used to the freedom of Android)</p><p><strong>What annoys me the most is NOBODY SEEMS TO CARE ABOUT THESE LIMITATIONS. I guess that’s what being “Apple user” is all about. Apple says no, you agree without any questioning. (iSheep)</strong></p><p>Anyway let’s start. Here are the workarounds I found for most issues.</p><h3>SONGS — Managing iTunes</h3><p>DISCLAIMER: This is for those who don’t buy songs from iTunes.</p><p>Let’s be honest no one buys songs. Like EVER! If you’re one of those keep on reading, others please leave i’m not interested in the ethical bla bla.</p><p>Or just buy the damn songs! But not me.</p><p>Managing your songs is a nightmare, even in Android, but since you have access to the folder structure in Android you can just make folders and organize your songs (One of the perks of Android).</p><p>For Apple users here’s the flexibility of Android:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ztc176FU2JwrpsvRCSSkEA.png" /><figcaption>In my S4 all my main songs were in my SD card, specifically in “Folder 1” and “Folder 2”</figcaption></figure><p>Now, when comes to Apple you need iTunes which requires all your meta data correctly (<strong>You can have songs but again this is about managing songs</strong>)</p><h3>Solution: TuneUp</h3><p><a href="http://www.tuneupmedia.com/">TuneUp | Fix Mislabeled Song Info, Add Album Art &amp; More</a></p><p>This is an life saving software! Not only did it help me with my whole song situation it actually made me clean up my whole song collection.</p><p>How you do it is simple, download and give access via iTunes. Then drag and drop your mp3 to iTunes and then iTunes to TuneUp. TuneUp will automatically find all your meta tag information:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lPtaCDwkoTmwcNTmn6ezsQ.png" /><figcaption>Example: How TuneUp works</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KGTZF5bZDriNNyn0Fw3peg.png" /><figcaption>Download the UK top 40 torrent and TuneUp will sort the albums</figcaption></figure><p>Pro tip: If TuneUp can’t find the album, Google (Wiki) and add the album name manually, this will help TuneUp to get the rest of tags.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yLjbPqfSbjDxxGB7Lf1iFw.png" /><figcaption>My iTunes library ^-^</figcaption></figure><h4><strong>Important things to remember</strong></h4><p>Don’t drag and drop all the songs to TuneUp. Do then gradually. Like if you have full albums which you have download, start with them. Then go song by song. (TuneUp says it can manage your whole collection but in my experience it messes up things here and there, i normally add 5 songs per time and update the tags)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*NGoGYtEB3yIbHq_0yZ7qbg.png" /><figcaption>WOLA! Everything is now organized perfectly!</figcaption></figure><h4><strong>Cracking and buying the software</strong></h4><p>I tried to crack this software many times it works sometimes but then TuneUp releases a new update and the crack no longer works (and have a fairly quick update schedule). So try your best to crack and tell me if you succeed.</p><p>When buying the software, download the trial version first and after some time they will popup a banner saying 50% off if you upgrade. That’s when i bought it. So don’t buy the software first, download the trial then get the offer.</p><h4>Freeware</h4><p>None of these works as good as TuneUp, but hey! try them and see for yourself.</p><p>MusicBrainz Picard<br>Jaikoz — for the geeks it’s purly written in Java!<br>MP3TAG<br>MediaMonkey — This is worst than iTunes. Don’t ever try this !!</p><p><em>This can be useful to any Android or iOS or anyone who likes there song collection.</em></p><h3>PICTURES</h3><p>Ready for this? 3 .. 2 .. 1 ..</p><blockquote><strong>You can’t copy pictures from your PC to iPhone!</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/1*LryIMhwOvB5tlwzxMpPEhQ.jpeg" /><figcaption>“Wait.. What??”</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/560/1*TXAih3Y3MPR8kDpCz2TtaA.png" /><figcaption>“I don’t know what this meme is called”</figcaption></figure><p>You can sync pictures via iTunes <strong>but </strong>the problem is you can’t do anything to those pictures. (You can’t delete, You can edit but that makes a copy of the picture which makes no sense to me. Why can’t i just copy my pictures?)</p><p>Again Apple claims it’s related to security. Apparently your pictures can contain viruses? (Google and see you can find some wild theories out there)</p><p>Let me explain how the Gallery works in iOS:</p><p>iPhone albums work differently. Your photos are only actually stored in only one album — the Camera Roll album. Even if you add a photo to another album, there is actually only one copy of that image file on your iPhone.</p><p>So the problem is organizing stuff. In Android you can select the pictures you want from Camera album and <strong>move</strong> them to a different album. So now if you go back to Camera album you have the rest of the <strong>unmoved</strong> pictures. But in iPhone you can’t do those since there is no move/copy/cut concepts in Gallery. It just references the pictures here and there.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*_POAcTO8yfUEa50Fxj2Exw.jpeg" /><figcaption>A picture stolen from the internet.. well simply because i can’t show you my Gallery :P</figcaption></figure><p>So back to my problem the</p><h4>only solution i found (which works so far is)</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*bqpfTPhALo8aAcoPKMQvhQ.png" /><figcaption>You have to buy this app</figcaption></figure><p>Now the reason I bought this app was it does what I want (95%)</p><h4>Ok now the technical stuff coming in… 3. 2. 1..</h4><p>What this fellow does is, it creates a local wifi network so your PC and your phones are in the same network.</p><p>What also works is you can upload all your pictures to DropBox or Google Drive and download them through iPhone (which i’m sure what everyone does) and this app does the same. It transfers all your picture from your PC to the “Camera roll” album of your phone. NOW you can do what ever you want with your pictures:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AN6tsEu4bX_VVwzgXw2hJQ.png" /><figcaption>You have upload picture to your excising albums</figcaption></figure><p>I’m still not comfortable with this approach but for the time being this helps my situation. I’m trying my best to get around stuff without jailbreaking. Which i’ll do soon…</p><p>(Even currently i’m transferring my screenshots using this app which I use in this article)</p><h3>PDF</h3><p>This was surprisingly easy! Just open the PDF in Safari (Share -&gt; open in Safari) and</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*1kvREUVNLnoX3-OVhanPfQ.png" /><figcaption>“Save PDF to iBooks”</figcaption></figure><h3>3rd PARTY FILE MNAGERS</h3><p>This was a dead end for me.</p><p>But i’ll tell you how to get that working also.</p><p>1 download a file browser</p><p>2 hook up the phone to PC and open iTunes</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*66jbf6h-g6o1HLLrHse08g.png" /><figcaption>You can add what ever the file type you want.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*Sree_T7KBgsqkdFhDHY4lg.png" /><figcaption>“Not bad eh ?”</figcaption></figure><p>But the problem is this, what ever the pictures or PDF or what ever you add here, iPhone has no idea about it. Meaning the pictures won’t show on your Gallery app. This is good for offline managing stuff like Microsoft Word and stuff but you’re better off using Google Drive or SkyDrive (or OneDrive what ever Microsoft uses.. ya that one)</p><h3><strong>SUMMARY</strong></h3><p>I hope this article helps someone out there, especially if you’re planing to switch to Apple but here’s the thing. I know according to this article iPhone is stupid but the reason i switch to Apple is still solid.</p><ul><li>I got 3 software updates (9.0.1, 9.0.2, 9.1)</li><li>Performance is snappy no lags what so ever</li><li>And day-to-day stuff (managing social media) is easy in iPhone</li></ul><p>What “Grinds my gears” is that every one of my Apple using friends have given up. Either they don’t listen to songs, or uses YouTube or there was this one guy (a long time Apple user), the requirement of ‘managing my gallery’ has never occurred to him. or something.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NPt-Cy8zkbR4E5eodcQM0Q.jpeg" /></figure><p>Maybe it’s my fault, my obsessive need to know how things work might work for Android but not for Apple.</p><p>Oh well…</p><p>Tell me what you think about this article. I’m sure hard core Android fans are over the moon after reading this article but hey, i still thinks iOS is better :)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=720f848d64af" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Galaxy S4 vs iPhone 6S camera test]]></title>
            <link>https://anujalamahewa.medium.com/galaxy-s4-vs-iphone-6s-camera-test-c334096164bc?source=rss-58afef076ad2------2</link>
            <guid isPermaLink="false">https://medium.com/p/c334096164bc</guid>
            <category><![CDATA[samsung]]></category>
            <category><![CDATA[iphone]]></category>
            <dc:creator><![CDATA[Anuja Lamahewa]]></dc:creator>
            <pubDate>Sun, 25 Oct 2015 12:45:42 GMT</pubDate>
            <atom:updated>2015-10-25T12:45:42.747Z</atom:updated>
            <content:encoded><![CDATA[<p>Finally had a free time to upload these.</p><p>A</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xpXqNIBt87Z84DAcInmbXQ.png" /><figcaption>S4 left — iPhone right</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9V4l4pd9tOhz93LmyF1zmQ.png" /><figcaption>S4 left — iPhone right</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H8IK7eZXhuc42p_6CMn1FA.png" /><figcaption>S4 left — iPhone right</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9fn5UWkEpQ3IOm1vLnQmAA.png" /><figcaption>S4 left — iPhone right</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c334096164bc" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>