Naive byte-range caching can waste gigabytes of origin bandwidth on small requests. We rebuilt our cache architecture in Pingora to coordinate partial fetches automatically, cutting R2 p99 latency in half.
We just shipped support for the ugliest part of HTTP: Vary
Vary support is now available in Cache Rules on every plan. You can normalize known negotiation headers, pass exact values through to the origin when those small differences matter, or bypass cache when the variation is too unpredictable.
Cross-Origin Storage API: Stop Downloading Twice
The Cross-Origin Storage API lets sites share one cached copy of a file (a JS library, font, or AI model) across origins by SHA-256 hash. Here's how it works.
Launching QuickDID - Fast, Open Handle Resolution for the AT Protocol - Nick's Blog
QuickDID is a high-performance, open source handle resolution service for the ATmosphere that serves as both public infrastructure at https://quickdid.smokesignal.tools and deployable software under MIT license. It offers flexible caching strategies (memory/Redis/SQLite), scales from single-instance to distributed deployments, and includes production features like rate limiting and proactive cache refresh. Currently a release candidate, it provides a drop-in alternative to Bluesky's resolver while giving developers full control over their handle resolution infrastructure.
turbopuffer: fast search on object storage
Inaugural blog post about the development of turbopuffer, a search engine that uses object storage and SSD caching for cost-effective, low latency search. This post describes into the motivation behind its creation, its unique architecture, and how it significantly reduces costs for large-scale vector searches. Discover how turbopuffer is transforming search infrastructure for companies like Cursor and Suno, offering a scalable and reliable solution.
quick_cache - Rust
Lightweight, high performance concurrent cache. It allows very fast access to the cached items with little overhead compared to a plain concurrent hash table. No allocations are ever performed unless the cache internal state table needs growing (which will eventually stabilize).
We're making Tailscale faster
Tailscale performance updates reduce memory use, increase throughput, and speed startup with multi-queue, writev, and netmap caching.
Optimizing Protocol Buffers in Go Applications
If you’ve ever wondered why your Go service is eating memory like it’s at an all-you-can-eat buffet, chances are you haven’t optimized your Protocol Buffers usage. I’ve been there, watching heap profiles with the kind of horror usually reserved for checking your bank account after a night out. But here’s the good news: Protocol Buffers in Go can be wickedly fast and memory-efficient when you know the tricks. Let me walk you through the optimization techniques that transformed my services from memory-hungry monsters into lean, mean, serialization machines.
Bunny Storage | The fastest edge storage accelerated by a CDN
Hop on the fastest object storage accelerated by a CDN. Automatically replicate your files worldwide to the edge to minimize the latency & maximize throughput.
Apache Iceberg: The Definitive Guide | Dremio
Learn everything about Apache Iceberg, the open table format for big data, in this definitive guide by Dremio. Discover best practices, architecture insights, and implementation strategies.
In fact, one of the leading local-first libraries, Zero, doesn't even fully support offline: zero.rocicorp.dev/docs/offline They only support offline reads, the same as Count Cachula! So I ask, if fast instant reads is the goal, isn't standard HTTP caching mechanisms the better way to go?
Offline
zero.rocicorp.dev@dholms.at can we fix that? iirc the rate limit headers RFC allows them on all responses? (Though obviously caching GET at CDN might skew things)
alice
i love services that are rate limited but also do not publish their rate limits in headers i'm looking at you plc.directory
