Re: crypto-ISSUE-23: Should CryptoOperations and/or Keys support Transferrable semantics?

On Mon, Aug 20, 2012 at 7:20 AM, Web Cryptography Working Group Issue
Tracker <sysbot+tracker@w3.org> wrote:
> crypto-ISSUE-23: Should CryptoOperations and/or Keys support Transferrable semantics?
>
> http://www.w3.org/2012/webcrypto/track/issues/23
>
> Raised by: Ryan Sleevi
> On product:
>
> Transferrable allows an object to be passed over a MessagePort, which allows it to be used with Web Workers. During initial discussions, it was suggested that CryptoOperations or Keys should have defined Transferrable semantics.
>
> The outstanding questions are:
> 1) Is there consensus to support Transferrable semantics?
> 2) If so, what are the use cases for Transferrables?
> 3) Because CryptoOperations represent objects with bound callbacks, what should the behaviour be for these Callbacks?
> 4) What happens if an object is Transferred in the midst of an operation?
> 5) If Key participates in the "structured clone algorithm", does it also need to be Transferrable?
>

This is similar to the clone() question, but specifically about
whether either object should support the "structured clone algorithm"
and/or be Transferrable. This is also closely related to ISSUE-8,
which is about describing key neutering.

Received on Monday, 20 August 2012 14:36:48 UTC