Parth Agrawal

f*ck it I am building clash royale in rust form scratch currently building the game env using bevy will post everyday my progress and journey documenting the whole precess in md files

time to revise and deep dive redis

Gemma 4: Google’s FREE OpenSource AI Powerhouse (Run It Locally) - The Big Shift: Google DeepMind just dropped Gemma 4 a new family of open-weights models bringing multimodal intelligence and advanced reasoning to everyone, for free. - "Thinking" for All: Every model includes a native reasoning mode, allowing step-by-step thinking to solve complex problems directly on your own hardware. - Four Powerful Sizes: The lineup includes E2B and E4B (optimized for mobile/laptops), a fast 26B-A4B (Mixture-of-Experts), and the top-tier 31B for heavy reasoning tasks. - Run it on a Laptop: The smaller models (E2B/E4B) are optimized for on-device usage, running on as little as 5GB RAM (4-bit). - Multimodal Intelligence: Built to handle text, images, and video natively across all models, with audio support in E2B and E4B. - Massive Context: Supports up to 128K tokens (small models) and up to 256K tokens (26B & 31B), enabling long-context reasoning. - Coding & Agent Ready: Major improvements in coding benchmarks, with native function-calling and system prompts for building autonomous agents locally. - Global Reasoning: Supports 140+ languages and uses hybrid attention to balance speed with deep long-context understanding. - Insane Local Speeds: With Unsloth Studio, you can reach up to 140 tokens/sec on high-end GPUs making local inference extremely fast. - Benchmarking Beast: The 31B model scores 85.2% on MMLU Pro and 89.2% on AIME 2026 (no tools), outperforming significantly larger previous models. (Gemma 4 is available now for free download, enabling developers to run and fine-tune powerful AI models without subscriptions.)

I built a Redis clone in Rust while back and then, for reasons that were absolutely not avoidable, gave it a TUI. check it out - [github.com/cotishq/Rustis]

@mahanot_dikshit I guess maine ye post dekha tha 😁

Things i got from reddit in last 3 days : - 500 karma - 40-50 compliments - 1 job offer

@kroy_3 Will write fs

@shankstwin Write a blog on these topics. Would love to read.

Since 1 week I am deep diving Postgres Here's what I covered : - MVCC — how Postgres avoids read locks - WAL — Write-Ahead Log, crash recovery, checkpoint - VACUUM and autovacuum — dead tuple cleanup - Connection pooling with PgBouncer — pool modes - Row-level security — tenant isolation at DB layer - Logical & streaming replication - Partitioning — range, list, hash, partition pruning - Full-text search — tsvector, tsquery, GIN indexes - JSONB with GIN index — flexible dynamic data - Window functions — ROW_NUMBER, LAG, LEAD, PARTITION BY - CTEs - common table expressions, recursive CTEs for trees - B-Tree — node structure, split/merge, height, range scans - Hash index — O(1) equality, not suitable for range - GiST — generalized search tree, geometric/range types - BRIN — block range, massive naturally-ordered tables - Partial indexes — WHERE clause, smaller, faster updates - Expression indexes — index on function result - Covering indexes — INCLUDE clause, index-only scans - EXPLAIN ANALYZE — reading query plans, actual vs estimated - N+1 query problem — identifying and fixing with JOIN Still learning covering more things I am really enjoying how things work and revising old concepts Will keep updating this list

@PhantomC0der Subah subah rust waah

starting the morning with some rust🦀🦀

@reddhairshanks Right 😎

Left or Right?

@adxtyahq fr i was using claude free version for some basic qna only 3 qs and over

claude code limits are becoming a real blocker you can’t build properly if you’re constantly hitting caps this is how you lose devs.

@LaniakeaG thanks for the suggestion I will dig into them fs

@shankstwin Solid list. One thing that’ll hit different in production: MVCC bloat. Long-running transactions hold back VACUUM from cleaning dead tuples, and table bloat creeps up silently until your queries slow to a crawl. Worth digging into pg_stat_user_tables and n_dead_tup early.

Back to the long sessions and deep focus—it feels great. Today’s PostgreSQL deep dive was all about Transactions & Concurrency: - ACID — atomicity, consistency, isolation, durability - Isolation levels — read committed to serializable - Concurrency anomalies — dirty read, phantom read, write skew - Deadlocks — detection, prevention, ordering - Two-phase commit (2PC) — distributed transactions - Saga pattern and Compensating Transactions today I slighly touched the concept of Distributed Systems ,felt interesting I am enjoying this fr

@LunarResearcher 😐💀

My girlfriend asked why I was smiling at 4 AM. Showed her the terminal. "What are all those green numbers?" $1,129. Made while she slept. "Doing what?" Nothing. Claude scanned 14,000 wallets, found 47 that never lose, built a bot that copies them. She watched for 10 seconds: +$3.87 captured +$6.42 captured +$12.71 captured "It just keeps going?" Every few seconds. New line. New money. "How much did you start with?" $300. Now $1,429. Eleven hours. Asleep. "What does it do?" Buys at $0.48. Sells at $0.52. Pockets $0.04. Who wins doesn't matter. "That's legal?" Citadel does this on NYSE daily. 400 engineers. I have one screen. Copy my bot here: @lunar" target="_blank" rel="nofollow noopener">kreo.app/@lunar She looked at the P&L curve. Never dips. Just climbs. "Can you make me one?" Setting hers up now. She still doesn't get how it works. The bot doesn't care.

We’re introducing Cursor 3. It is simpler, more powerful, and built for a world where all code is written by agents, while keeping the depth of a development environment.

@gurshabad90 happy birthday paaji 🥳🥳🥳

Balloons 🎈😁

We just released Gemma 4 — our most intelligent open models to date. Built from the same world-class research as Gemini 3, Gemma 4 brings breakthrough intelligence directly to your own hardware for advanced reasoning and agentic workflows. Released under a commercially permissive Apache 2.0 license so anyone can build powerful AI tools. 🧵↓

Google AI Pro Needs More AI Power, Not More Storage Google AI Pro moving from 2TB → 5TB is nice but honestly, thats not what most users signed up for. Most people bought Google AI Pro for AI usage not storage. What Google AI Pro already does well: - 5TB cloud storage (great, but not the main reason people buy it) - Access to Gemini CLI and tools like Jules - 1000 AI credits for usage - NotebookLM, etc - Antigravity (but low quota and still feels unfinished) We don’t need more storage. We need a plan focused purely on AI. Give us: - Higher AI usage limits - Better Antigravity quota - A more polished experience A dedicated “AI-only” plan would make way more sense for actual users.

@buildwithsid This is how linkedin works bro 😭

someone made that hand gesture camera thing again and got 1k+ likes on Linkedin meanwhile I posted my projects and got 7 💀

someone just dropped a way to get insanely clean UI out of claude code the problem till now: - no context of your codebase - constant switching between tools - UI that doesn’t actually fit your project this fixes that, now claude can: - design inside your workflow - understand your system - generate UI that actually fits try it here - aidesigner.ai/docs/claude-co…

After 1 month u realize u submitted wrong notebook for the assignment