Re: [w3ctag/spec-reviews] IndexedDB 2.0 features review (#153)

> ...since binary keys expect the underlying type information to be discarded when stored...

There may be developer confusion that e.g. a Uint8Array and Float32Array consisting of the same backing bytes have the same value as a key. (This was touched on in the linked discussion). This pattern shows up in a few other places in the platform where APIs consume binary data; we eventually converged on the conventions specified in WebIDL's "get a copy of the bytes..." and the BufferSource abstract type. So the platform is consistent, but again there is the possibility of developer confusion.

> ...use case behind allowing renaming...

Most of the discussion around store/index renaming took place offline, and I failed to capture the intents there. (This also included guidance from @annevk on using setter rather than e.g. setName or anything awkard like that.)

This blog post by @bevis-tseng covers the use cases pretty well: https://hacks.mozilla.org/2016/10/whats-new-in-indexeddb-2-0/

> ...key generators...

Nothing (intentionally) new with key generators in 2.0 vs. 1.0, although we have identified a spec gap and interop issues with edge cases, discussed at https://github.com/w3c/IndexedDB/issues/147 - more eyes on everything welcome!


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/w3ctag/spec-reviews/issues/153#issuecomment-281434171