Coffee With Developers Aug 31, 2026

What is Cross-Origin Storage - Thomas Steiner

Thomas Steiner

Thomas Steiner proves shared browser caching isn't dead. Discover how the Cross-Origin Storage API securely shares massive web AI models across domains without triggering privacy-compromising timing attacks.

Pause
Mute Enter Fullscreen
#1 about 3 min

Overcoming historical web caching and privacy challenges

How partitioned browser caches broke shared content delivery networks to prevent timing attacks.

#2 about 3 min

Solving storage constraints for web AI and fonts

Why large assets like local AI models and comprehensive emoji fonts require a new caching mechanism.

#3 about 2 min

Addressing shared web resources by content hash

Using content-addressed storage to verify and share resources regardless of their original download URL.

#4 about 3 min

Managing sharing levels with the imperative API

How to use JavaScript to query the cache and configure global, same-site, or origin-specific sharing.

#5 about 2 min

Opting into shared caching with declarative HTML attributes

Adding specific attributes to script and link tags enables seamless integration with cross-origin storage.

#6 about 4 min

Preventing malware distribution through decentralized hash verification

Validating incoming bytes against known hashes stops compromised domains from serving malicious code.

#7 about 3 min

Understanding underlying browser storage implementations and tooling

How browser vendors might adapt existing mechanisms like IndexedDB or OPFS to store cached blobs.

#8 about 4 min

Mitigating timing attacks and managing storage quotas

Protecting user privacy by enforcing popularity thresholds and probe limits on cached resources.

#9 about 3 min

Expanding ecosystem adoption across WebAssembly and AI frameworks

How partners use the API to share complex dependencies like Skia rendering engines and compiled database libraries.

#10 about 2 min

Testing the proposed API using browser extension polyfills

Developers can experiment with the cross-origin storage proposal immediately using offscreen page service workers.

#11 about 3 min

Integrating storage across CSS and Fetch API endpoints

Modifying standard request options allows developers to cache assets through font-face declarations and network requests.

#12 about 6 min

Providing usage metrics to support browser vendor implementation

Expressing developer interest with concrete metrics helps prioritize the API standardization process.

#13 about 2 min

Integrating shared caching with modern web framework bundlers

Using plugins for tools like Vite enables frameworks to optimize resource delivery natively.

Matching moments

2:53 min

Solving cross-origin model size and native performance limitations

Christian Liebel Christian Liebel · World Congress 2025

2:44 min

Introduction to web acceleration and caching paradigms

Thijs Feryn Thijs Feryn · World Congress 2025

2:57 min

Managing local storage footprints for progressive browser AI systems

Chris Heilmann +2 · LIVE

2:46 min

Querying cold storage directly from the browser

Thorsten Riess Thorsten Riess · Europe 2026 Virtual

6:38 min

Improving asset caching through localized font deployments

Chris Heilmann +2 · LIVE

4:54 min

The origins and design principles of WebAssembly

Maurice Brinkmann · World Congress 2023

Upcoming sessions on this topic

Open session

World Congress 2026 North America

September 24, 2026 · 10:20–10:50

Stage 1

The Browser Is Becoming an AI Runtime

Ajit Varma, Will Bryk

Ajit Varma
Will Bryk
Open session

World Congress 2026 North America

September 25, 2026 · 16:10–16:40

Stage 4

Speed up your CI/CD pipelines by caching build & runtime artifacts

Thijs Feryn

Technical Evangelist at Varnish Software

Thijs Feryn
Open session

World Congress 2026 North America

September 25, 2026 · 15:30–16:00

Stage 6

Small LLM in your Browser: Huge Opportunities for Web Applications

Daniel Ostrovsky

UI/UX Architect at Payoneer | AI Architect | Full Cycle Development Expert | Public Speaker | Open Source Contributor |

Daniel Ostrovsky
Open session

World Congress 2026 North America

September 24, 2026 · 14:10–14:40

Stage 1

Anatomy of an AI Request: Where Latency and Cost Are Really Born

Dan Fu

VP of Kernels at Together AI

Dan Fu
Open session

World Congress 2026 North America

September 25, 2026 · 14:10–14:40

Stage 4

Headroom: A Context Optimization Layer for LLM Applications

Tejas Chopra

Senior Software Engineer at Netflix

Tejas Chopra
Open session

World Congress 2026 North America

September 24, 2026 · 15:30–16:00

Stage 2

KV Cache Is Not About Speed: It's About Surviving Inference Costs

David vonThenen

AI/ML Leader | Keynote Speaker | OSS Engineer & Developer Advocate | Agentic AI, Deep Learning, Production AI | Python, Go, C++

David vonThenen