Fetch: Integration with the Cache API

[james-pre]

@billiegoose:

I'm not actually working on this yet (I'm still using IndexedDB in the Service Worker), but I think it would be cool so I'll write this idea down anyway.

It's still experimental, but the Cache API is very cool:

• it's a key / value store
• for arbitrary files (no need to serialize/deserialize)
• supports streaming
• no fixed size limits (limits are determined by the quota API which is still in development)
• caches global available in both Window and Service Worker scopes
• supports quickly downloading and caching a batch of files for offline access via Cache.addAll()

@jvilk:

Sounds nice. I would call it CacheStorageFS instead, since I would like to have some sort of generic caching file system implemented in the future (jvilk/BrowserFS#95).

jvilk/BrowserFS#191

[james-pre]

Looking into the Cache API, it can only cache Requests/Responses. That means it is limited to fetch requests.