Roberto Venditti’s Post

🍦 @fizzwiz/vanilla v0.0.0-dev.3 is here! We’re excited to release v0.0.0-dev.3 — bringing declarative exploration and enhanced object navigation to your JSON workflows. ✨ Highlights: • search() enables fluent traversal of nested objects. • delete() and select() make removing or filtering entries effortless. 🧠 Learn more: https://lnkd.in/eawSVFnW 📦 Grab it on npm: https://lnkd.in/eqg6EvCy — @fizzwiz ✨

Law of Leaky Abstractions - JSON is not an object. It is a string.  - S3 is not a file system. It is a key-value store for blobs. - Notion renders markdown. You can’t write markdown in Notion.

Does MoreLINQ Zip Outperform LINQ Zip? Let's Benchmark Them! A duel between two methods for zipping collections in CSharp! But a challenger approaches: Our very own naive implementation of LINQ Zip! In this video, I'll walk through the BenchmarkDotNet benchmarks for comparing the LINQ zip method and MoreLINQ ZipShortest and ZipLongest methods. We'll even see how my naive implementation does against these! Buckle up and get ready to see how optimized my code is! Spoiler Alert: It's not. More Spoiler Alert: You'll still be surprised. Watch here: https://lnkd.in/gY4nmEbe

The point of JSON, objects remember how they look like. In TOON (token oriented object notation) objects have no idea how they look like. TOON works for very simple cases, but in those cases you probably don’t need TOON or JSON, or any other “language”. Am I missing something?

 1️⃣Binary Search Beyond LeetCode 💭 “We all know binary search finds an element fast. But in real systems, it helps you design smarter logic.” Example: Load balancing (finding threshold load) Feature rollout (finding smallest % of users with an issue) 🎯 Takeaway: Binary Search = Decision search, not just array index search.

What is TOON Format? TOON is a lightweight, human-readable way to display structured data without the heavy syntax of JSON. It’s not a real file format—just a clean, compressed visual style developers use in posts and presentations. JSON → machine-friendly TOON → human-friendly

🍦 @fizzwiz/vanilla v0.0.0-dev.4 is here! We’re excited to release v0.0.0-dev.4! This version introduces path-aware navigation: • Each ObjNavigator now tracks the step from its parent. • The 𝗉at𝗁() method lets you compute logical or structural depth in nested objects. 🧠 Learn more: https://lnkd.in/eBWg9Pe5 📦 Grab it on npm: https://lnkd.in/eqg6EvCy — @fizzwiz ✨

Your RAG app is useless. You probably use plain semantic search to retrieve relevant documents. However, neglecting knowledge graphs is your biggest mistake. With 𝗰𝗼𝗴𝗻𝗲𝗲, you can implement Knowledge Graphs in just five lines of code. 𝗰𝗼𝗴𝗻𝗲𝗲 is an open-source library that gives your agents dynamic memory. Here's why you should use it: ✅ Load data into graph and vector databases such as Qdrant. ✅ Interconnect and retrieve your past conversations, documents, images and audio transcriptions ✅ Ingesting data from over 30 data sources With 𝗰𝗼𝗴𝗻𝗲𝗲, you won't experience any downgrade compared to your traditional retrieval method. 🔗 Link to repo: github(dot)com/topoteretes/cognee

Can TOON(Token-Oriented Object Notation) replace JSON for LLM ? TOON conveys the same information with fewer tokens 💸 Token-efficient: typically 30–60% fewer tokens than JSON 🤿 LLM-friendly guardrails: explicit lengths and field lists help models validate output 🍱 Minimal syntax: removes redundant punctuation (braces, brackets, most quotes) 📐 Indentation-based structure: replaces braces with whitespace for better readability 🧺 Tabular arrays: declare keys once, then stream rows without repetition https://lnkd.in/dgSVdMXN

Day 18 of #100DaysOfDSA Today I learned about Tree Traversal Algorithms — focusing on Depth-First Search (DFS) and its types: Inorder, Preorder, and Postorder. 🌳 What I Learned: DFS (Depth-First Search): Goes deep into each branch before moving to the next — useful for exploring or analyzing the structure of trees. Inorder: Left → Root → Right (gives sorted order in BST) Preorder: Root → Left → Right (used for creating or copying trees) Postorder: Left → Right → Root (used for deleting or freeing nodes) BFS (Breadth-First Search): Explores level by level and is often used for finding the shortest path, such as in social networks or maps. 💡 Takeaway: DFS dives deep, BFS spreads wide — both are essential for understanding how data connects and flows. 🔗 Here’s my implementation on GitHub: https://lnkd.in/gYAa7z6P #100DaysOfCode #DSA #JavaScript #TreeTraversal #DFS #BFS #LearningInPublic