SSachin Rajput1h ago · 12 min readPrefill vs Decode: Two Phases, Two BottlenecksIn Part 1 the Word Factory read your whole question in one go, then typed the answer one token at a time. Reading was cheap per token. Typing was slow. In Part 2 we counted the math for one token, andsachinrajput.hashnode.dev
WWaseem Anjum4h ago · 18 min readMongoDB vs Postgres JSONB in 2026: Inserts, Lookups, Updates and Storage, BenchmarkedMost "MongoDB vs Postgres JSONB" debates are fought with benchmarks from 2014 and opinions from Twitter. I run both in my day job and side projects, so I put the current versions, MongoDB 9.0.4 (releafabwaseem.hashnode.dev
CCruxtab Technologies Pvt. Ltd.4h ago · 3 min readDesigning Safe and Efficient S3 Deletion for Backend ApplicationsDeleting files from Amazon S3 is straightforward when there is only one known object. The engineering challenge grows when an application needs to delete many files associated with form submissions. Acruxtab-technologies.hashnode.dev
ZZyVOP4h ago · 5 min readCaching Strategies Compared: Cache-Aside, Write-Through, and Write-BehindEvery cache forces you to answer one question: when data changes, who updates the cache, and when? Reads are the easy half. The write path is where caches go stale, lose data, or slow your app down. Tblog.zyvop.com
MMagnus9h ago · 12 min readWhy AI Agent Cache Hit Rates Collapse: GPT, Claude, and Auditable GatewaysA source-backed guide to third-party agent cache failures, GPT and Claude prompt-caching mechanics, reproducible hit-rate measurement, and Modelflare's public compensation boundary. AI agent caching imodelflare.hashnode.dev
YYunus Emre Vurgun5h ago · 10 min readBuy the Bottleneck, Not the Spec Sheet: A Workload-First Method for Sizing a Workstation, Server, or HomelabMost hardware regret has the same shape. Someone buys the part with the biggest number on the box (more cores, a newer GPU, a faster SSD) and the machine still feels slow at the one job they bought ityunusemrevurgun.hashnode.dev
TToolsy Hub8h ago · 3 min readHow to compress an image to an exact KB limit in the browser (canvas, binary search, and when to stop lowering quality)Upload forms for visas, exams and job portals often say "photo must be under 100 KB". A phone photo is 3 to 5 MB, and the usual advice is to drag a quality slider until the number looks right. You cantoolsyhub.hashnode.dev
DDominik Groński13h ago · 2 min readSelf-host your Draco decoder - the 600ms CDN tax on Three.js configuratorsDraco compression is the difference between a 90MB glTF export and a 3MB one. For CAD-heavy product models it is not optional. But there is a trap in the standard setup that quietly costs you half a sgrodev.hashnode.dev
NNitesh Kumar Tudu20h ago · 7 min readCloudflare Saved 100 TB of RAM With High-School Statistics. Here's the Lesson for Your Node BackendLast week Cloudflare published a post about cutting roughly 100 TB of RAM out of their internal load balancer. They didn't do it with a rewrite or new hardware. They counted how many hash points their2EAblog.thecodingant.in
AAlki20h ago · 7 min readVideo on scroll in Elementor — why scrubbing an MP4 stutters (and what to do instead)Originally published at: https://beta.scrollsequence.com/blog/video-on-scroll-why-scrubbing-stutters-mp4-vs-image-sequence Tutorials like this one from Nicolai Palmkvist that scrub a video with GSAP Sscrollsequence.hashnode.dev
