Ankit Kumar’s Post

I used Claude Code in two ways: normal agent mode in CLI and autonomous mode in Docker Container (with bypass-permissions on). Anthropic came out with sandbox mode, supposedly to bridge the gap between the two methods. At first, I never liked the mode because it kept stopping at random times because of some permission issues. Yesterday, I encountered it again, was determined to fix it. It was "bwrap: loopback: Failed RTM_NEWADDR: Operation not permitted", which was a problem of the sandbox failing to set up its network namespace. I found these two articles on how to fix the issue. https://lnkd.in/gTrTV66A https://lnkd.in/gi8NbVYX Afterward, Claude Sandbox mode worked perfectly. It was able to utilize a Python script to read a docx file, since I used a lot of docx file with Claude.

Just released version 10.8.3 of passagemath, the comprehensive #OpenSource #Mathematics system in #Python, the pip-installable modularized portable fork of SageMath. It is available: - on PyPI, providing source distributions and binary wheels for Linux and macOS [full functionality of #SageMath, GAP, Macaulay2, and more] and Windows [partial functionality], - on the prefix.dev emscripten-forge channels, shipping #WebAssembly packages [partial functionality] for serverless in-browser use with Project Jupyter's JupyterLite and Notebook.link, - in the MSYS2 software distribution for Windows [partial functionality]. The release 10.8.3 updates the #CAS Maxima to version 5.49.0, updates GNU MPC (multi-precision complex numbers) to 1.4.0 and OpenBLAS to 0.3.32, and includes various bugfixes backported from #SageMath development. Developers who prefer building passagemath from source will benefit from the initial support for the clang 21 toolchain and autoconf 2.73. passagemath's updated version of the Sage distribution now ships the marimo Jupyter Lab extension for interoperability. It also brings various updates, including CVXPY to version 1.8.2 and NumPy to version 2.4.3. https://lnkd.in/gUmAmzQ8

Just released version 10.8.4 of passagemath, the comprehensive #OpenSource #Mathematics system in #Python, the pip-installable modularized portable fork of SageMath. It is available: - on #PyPI, providing source distributions and binary wheels for Linux and macOS [full functionality of #SageMath, GAP, Macaulay2, PARI/GP, and more] and Windows [partial functionality], - on the prefix.dev emscripten-forge channels, shipping #WebAssembly packages [partial functionality] for serverless in-browser use with Project Jupyter's JupyterLite and Notebook.link, - in the #MSYS2 software distribution for Windows [partial functionality]. In passagemath-10.8.4, thanks to work by Samuel Chen, rich output is now also supported in Jupyter notebooks with standard Python kernels, such as on Google Colab or in JupyterLite (XPython). Example JupyterLite notebook for Algebraic Combinatorics, now with some plots: https://lnkd.in/gveZVfKE https://lnkd.in/g2kXE4Vd

Me before today: "Yeah, I know what ThreadPool is... theoretically." Me after today: Actually implements background workers, fixes race conditions with locks, and makes code run 3x faster. The "Aha!" moment: running multiple tasks simultaneously vs. one by one and watching the execution time drop from 6 seconds to 2 seconds.  Level 3 task with Codveda done! #CodvedaAchievements! Finally put those OS theory concepts (Linux daemons, who? 😂) into real C# code. Biggest wins: Async/await = non-blocking magic ThreadPool = reusing workers like a boss lock = stopping threads from fighting over data Simulated daemons in Windows (feeling powerful) Theory → Practice = Growth 📈 Anyone else remember their first "aha" moment with async programming? #Codveda #CodvedaProjects #CSharp #NeverStopLearning #threading

🚀 Day 6/10 — Optimization Series Python Virtual Environments (Why They Matter) 👉 Basics are done. 👉 Now we move from working code → optimized code. You install a package… Everything works fine… Then suddenly 😐 👉 Another project breaks 🔹 The Problem 👉 Different projects need different versions 👉 Global installation = conflicts 🔹 What is a Virtual Environment? 👉 A separate isolated space for your project Each project can have: its own packages its own versions 🔹 Create Virtual Environment python -m venv myenv 🔹 Activate Windows: myenv\Scripts\activate Mac/Linux: source myenv/bin/activate 🔹 Install Packages pip install pandas 👉 Installed only inside this environment 🔹 Why This Matters No version conflicts Clean project setup Easy collaboration 🔹 Real-World Use 👉 Multiple projects 👉 Team development 👉 Production environments 💡 Quick Summary Virtual environment = isolated workspace 💡 Something to remember If everything breaks after installing one package… you’re not using virtual environments. #Python #DataEngineering #LearningInPublic #TechLearning

Avalonia has released an initial preview of an Avalonia backend for .NET MAUI. The framework uses .NET 11 (itself in preview), and will mean developers will be able to deploy .NET MAUI apps to Linux and WebAssembly.  Avalonia started life as a cross platform .NET UI framework with XAML, data binding, and lookless or templated controls. Avalonia's initial support for developing XAML-based applications for Windows, Mac and Linux has been widened to become an open source UI framework for building desktop, mobile, web and embedded applications using a .NET single codebase.

I built a free tool that saved my data — and it might save yours too. A while back, my hard drive started failing. I tried to copy files off it before it died completely. Windows Explorer froze. Then crashed. Then the system locked up entirely. By the time I rebooted, the drive was worse. The problem? Explorer retries every read error indefinitely — it just hangs, burning precious seconds on a dying disk. So I built SwiftCopy — a fault-tolerant file recovery tool for Windows. It wraps robocopy in a clean dual-pane GUI and applies one simple rule: skip errors, keep moving. Zero retries. Skip locked files. Get your data out before the drive gets worse. Features: - One-click preset for failing drive recovery - Dual-pane browser (Total Commander–style) - F5 to copy, F6 to move - Real-time progress, speed, and log output - Standalone .exe — no Python, no install It's free, open source, and available now on GitHub. Download: https://lnkd.in/gUVtgUPT If you've ever lost data to a freezing Windows copy dialog — this is for you. Share it with anyone who might need it. #OpenSource #Windows #DataRecovery #Python #Qt #SideProject #SoftwareDevelopment

I built a free tool that saved my data — and it might save yours too. A while back, my hard drive started failing. I tried to copy files off it before it died completely. Windows Explorer froze. Then crashed. Then the system locked up entirely. By the time I rebooted, the drive was worse. The problem? Explorer retries every read error indefinitely — it just hangs, burning precious seconds on a dying disk. So I built SwiftCopy — a fault-tolerant file recovery tool for Windows. It wraps robocopy in a clean dual-pane GUI and applies one simple rule: skip errors, keep moving. Zero retries. Skip locked files. Get your data out before the drive gets worse. Features: - One-click preset for failing drive recovery - Dual-pane browser (Total Commander–style) - F5 to copy, F6 to move - Real-time progress, speed, and log output - Standalone .exe — no Python, no install It's free, open source, and available now on GitHub. Download: https://lnkd.in/g3wr3jmD If you've ever lost data to a freezing Windows copy dialog — this is for you. Share it with anyone who might need it. #OpenSource #Windows #DataRecovery #Python #Qt #SideProject #SoftwareDevelopment

🚀 Compiler Construction Project – Demo Video Released! I’m excited to share my recent project in Compiler Construction, where I implemented a working Lexical Analyzer (Scanner) as part of a mini compiler design system. 💡 Key work in this project includes: Regular Expressions for token definitions Transition diagrams for Identifiers & Numbers Token recognition (keywords, operators, punctuation) Error handling in lexical analysis Generation of token output from a Mini C++ program 🎥 Demo Video: The video demonstrates how the scanner processes source code step-by-step and generates tokens successfully. 📁 Project Repository: 👉 https://lnkd.in/duHXf5eT This project really helped me understand how a compiler works at its initial stage and how high-level code is converted into meaningful tokens. 🔧 Tools Used: Flex / C++ / Ubuntu (Linux) I’d really appreciate your feedback and suggestions! #CompilerConstruction #LexicalAnalysis #Flex #CPlusPlus #GitHub #ComputerScience #Programming #SoftwareEngineering #StudentProject

Pairing GDAL for a geospatial workflow in Python today. 💻 Mac vs Windows (A Love Letter to GDAL… and a Slight Cry for Help 😄) If you’ve ever tried installing GDAL on Windows, you know the struggle. If you’ve ever tried installing GDAL on Windows, you already know… it’s not just installation — it’s a journey. You start with: “pip install gdal” …and suddenly you’re: Installing dependencies you’ve never heard of. Fixing PATH variables like a detective Questioning your life choices. Then I finally typed one line in PowerShell: ✨ BOOM. It works. That moment? Pure joy. Like unlocking a hidden level in life. #python #rosettastone #gisworkflows