Dev interest in using Storage Buckets API

WICG / storage-buckets

API proposal for managing multiple storage buckets

https://wicg.github.io/storage-buckets/explainer

Other

44 stars 20 forks source link

Dev interest in using Storage Buckets API #92

Open evanstade opened 1 year ago

evanstade commented 1 year ago

Hi web developers,

We would like to gauge interest in using the Storage Buckets API, as best described in the developer blog.

If you are interested in potentially using this new API, please reply to this thread, including your affiliation to a company or organisation or app, if any.

We plan to open it up to origin trial soon, so bonus points if you are interested in trying it out early.

Thanks!

carlopi commented 1 year ago

I will be interested in experimenting with this. Carlo Piovesan, DuckDB Labs, project is DuckDB-Wasm (https://github.com/duckdb/duckdb-wasm or https://shell.duckdb.org).

ayuishii commented 1 year ago

Storage Buckets API is ready for Origin Trial on Chrome starting M115-118. You can try the feature out on your website by registering and using the token provided. We’d greatly appreciate any feedback on the API. Thanks!

Origin Trial Registration: https://developer.chrome.com/origintrials/#/view_trial/2674012278751232001

Local testing available from M110: chrome://flags/#storage-buckets

tomayac commented 1 year ago

See https://github.com/whatwg/fs/issues/145 for a potential use case for Storage Buckets.

asakusuma commented 11 months ago

We (LinkedIn) are interested in using buckets + Clear-Site-Data to allow us to cleanly and reliably evict a service worker and any associated persistent data, without also logging out the user and destroying other persistent data. The ability to use Clear-Site-Data on a service worker without collateral damage will make it much easier to sell service worker experimentation internally, since there is some hesitation to use service workers because of the larger risks of bug impact in service worker code.

tepark-goog commented 11 months ago

YouTube is interested in the benefits that buckets provide, namely:

Quota policy enables better heuristics around how much we store for our networkless logging solution.
Per-bucket storage estimates allow for improved UX around downloaded video storage availability.
As the API matures, we are hoping to see some general IndexedDB performance gains. Examples:
- Some policies are immutable, thus optimizations can be made on bucket creation (e.g. different underlying storage structure for relaxed vs strict).
- DB operations in one bucket can be fully independent (process-wise) of DB operations in another bucket.

Sora2455 commented 8 months ago

Anecdotally, one use case I have in mind concerns JWTs. Currently, those are cached in client-side storage and manually cleared when the front-end detects a logout. If they were stored in a storage bucket, I would be able to use Clear-Site-Data to clear specifically the JWT bucket from the server-side.