DEV Community: Vichea Nath

🚀 Keeping Legacy Laravel Projects Alive in the PHP 8 Era

Vichea Nath — Fri, 23 Oct 2026 03:24:55 +0000

I’ve been using CRUDBooster since 2018, and over the years, I’ve built and maintained projects that rely on it. It has been a valuable tool for rapidly building Laravel applications, especially for admin panels and CRUD functionality.

Since then, my professional development journey has moved toward C# and .NET, which I now primarily use for my software engineering work. However, I still maintain side projects built with PHP, so I haven't completely left the PHP ecosystem behind.

Many applications built before PHP 8 are still running in production today. Upgrading these legacy projects to newer PHP and Laravel versions can be challenging, especially when the original packages are no longer actively maintained or newer versions introduce breaking changes.

That's one reason I'm continuing to work on my CRUDBooster Contrib fork. Having used CRUDBooster for so many years, I'd like to help developers who still depend on it maintain and extend their existing applications rather than rewrite everything from scratch.

🔗 GitHub: https://github.com/vicheanath/crudbooster-contrib

I'd love to hear from other developers:

Have you been using CRUDBooster since its earlier days?
Are you still maintaining Laravel projects built before PHP 8?
Have you encountered compatibility or security challenges with older CRUDBooster applications?
Are you still using PHP for side projects even though your primary stack has changed?

It's great to see how our technology stacks evolve over time, but some of our older projects still deserve attention. Let's share our experiences and help keep these projects running safely and reliably.

🧠 Vision AI Labeling Studio: The Open-Source Powerhouse for Image & Video Annotation

Vichea Nath — Tue, 20 May 2025 03:32:31 +0000

In the rapidly evolving world of machine learning and computer vision, high-quality labeled datasets are the cornerstone of successful AI models. Whether you're training an object detector, OCR system, or segmentation network, annotation tools play a vital role. Among the many options available, a new open-source tool stands out: Vision AI Labeling Studio.

Built with flexibility, power, and developer freedom in mind, Vision AI Labeling Studio offers an impressive set of features that rival—and in many ways exceed—those of commercial and closed-source alternatives.

🚀 Why Vision AI Labeling Studio?

✅ Feature-Rich & 100% Open Source

Unlike most commercial tools, Vision AI Labeling Studio is completely open-source, allowing full customization and local deployment without vendor lock-in or subscription fees. It's designed for individuals, teams, and enterprises looking for maximum control over their data and infrastructure.

🛣️ Feature Roadmap Comparison

Vision AI Labeling Studio is rapidly evolving. Here’s how its roadmap of planned and in-progress features stacks up against popular alternatives:

Feature	Vision AI Labeling Studio	Label Studio	CVAT	MakeSense.ai	Supervisely
✏️ Free Drawing (Lasso Tool)	✅	❌	✅	❌	✅
💾 Offline Project Storage	✅ (IndexedDB / SQLite)	❌	✅	✅	❌
🖼️ Multi-Image Projects	✅	✅	✅	✅	✅
🖥️ Desktop App (Electron)	✅	❌	❌	❌	❌
🤖 AI YOLOv8 Auto-Detection	🔄 In Progress	❌	Limited	❌	✅
📤 Export Multiple Formats	✅ (COCO, YOLO, Pascal VOC, JSON)	✅	✅	✅	✅
🏷️ Multi-Class Annotation	✅	✅	✅	✅	✅
🎥 Video Frame-by-Frame Annotation	✅	❌	✅	❌	✅
🖼️ Image Segmentation (Polygon)	✅	✅	✅	✅	✅
🖼️ Text Annotation (OCR)	✅	✅	✅	❌	✅
👥 Collaborative Labeling (Team Mode)	✅ (Self-hosted)	✅	✅	❌	✅
☁️ Cloud Storage Integration	✅ (S3, GCS, Azure)	✅	❌	❌	✅

🔧 Built for Real-World AI Workflows

Whether you're labeling medical scans, training OCR models, or building smart surveillance systems, Vision AI Labeling Studio is built to fit directly into your workflow:

Works Offline: No internet? No problem. Projects are stored in IndexedDB (Web) or SQLite (Electron).
Modular Architecture: Easily extend it with plugins or APIs.
Electron Desktop App: Seamless offline annotation and AI-assisted labeling.
Cloud-Ready: Use with Amazon S3, Google Cloud Storage, or Azure for scalable projects and team collaboration.

🧠 Smart AI Integration

One standout feature currently in development is YOLOv8-powered auto-labeling. This integration will allow users to kickstart labeling with pre-detected bounding boxes, dramatically speeding up the annotation process for object detection projects.

💬 Community and Customization

Vision AI Labeling Studio invites contributions and feature requests from the community. Being open-source, it’s not just a tool—you can shape it to your needs.

GitHub: https://github.com/vailabel/vailabel-studio
Frameworks: React (Vite), Electron, Python (FastAPI for cloud)

🆚 Final Thoughts

While commercial tools like Supervisely or Labelbox offer robust cloud ecosystems, they often come with usage limits or hefty pricing. Vision AI Labeling Studio breaks that barrier—offering professional-grade tools for free.

If you're a data scientist, ML engineer, or a small team looking for a powerful, flexible, and self-hosted annotation platform, Vision AI Labeling Studio is worth serious consideration.

🧠 Vision AI Label Studio – Build Better Datasets with Open Source Image Labeling (with AI!)

Vichea Nath — Sat, 17 May 2025 21:49:41 +0000

If you're tired of slow or clunky annotation tools, I built Vision AI Label Studio, a free and open-source labeling tool for computer vision datasets.

🔧 Features:

🎨 Draw bounding boxes, polygons, freehand tools
⚡ AI Auto-Labeling with YOLOv8
🖥️ Desktop + Web apps
🌓 Light/Dark mode
🗃️ Exports: COCO, YOLO, VOC, Simple JSON

📴 Works 100% offline, built with React + Dexie.js

This tool is ideal for AI developers, researchers, and data labeling workflows.
Check it out 👇
👉 https://github.com/vailabel/vailabel-studio

Would love feedback, contributions, and stars 🌟!

Get digitalocean 100$ freee

Vichea Nath — Tue, 22 Dec 2020 13:46:40 +0000

You get 100$ free on Degitalocean Hosting by Click this Link
https://m.do.co/c/a39a711827b3