W3C home > Mailing lists > Public > public-script-coord@w3.org > July to September 2013

Re: Structured clones

From: Anne van Kesteren <annevk@annevk.nl>
Date: Wed, 4 Sep 2013 21:18:47 +0100
Message-ID: <CADnb78hbo68zB7PoXBb1d1kWzN2PFSpNw-vkDFfE7peLaeTAxw@mail.gmail.com>
To: "Mark S. Miller" <erights@google.com>
Cc: es-discuss <es-discuss@mozilla.org>, "public-script-coord@w3.org" <public-script-coord@w3.org>
On Thu, Apr 11, 2013 at 11:19 AM, Anne van Kesteren <annevk@annevk.nl> wrote:
> On Wed, Apr 10, 2013 at 3:27 PM, Mark S. Miller <erights@google.com> wrote:
>> For those new to these lists,
>> <http://lists.w3.org/Archives/Public/public-script-coord/2011JulSep/0055.html>
>> may provide some context. The structured clone algorithm has changed since
>> then. But looking this over, I think most of these points are still
>> relevant. See also the rest of the thread starting at this message.
>
> With regards to #4 in that message, Blob/File/FileList are readonly,
> so there is no writing to the same object.
>
> Given that workers and indexed DB depend on this feature it is not
> going away. We could rebrand it as marshalling as suggested in that
> thread I suppose, but I was mostly interested if this was something we
> should be able to express in JavaScript as well.
>
> There's also a concept of transferable objects for objects to which
> you can write, such as ArrayBuffer.

This has come up again:
http://lists.w3.org/Archives/Public/public-whatwg-archive/2013Sep/0023.html

This concept will keep growing and is used by multiple APIs these
days. I don't really have any concrete proposals, but it seems like
something we should solve at the language level and not monkey patch
on top of it. It's probably late again as with ArrayBuffer to design
something from the ground up, but maybe there's still salvageable
bits.


-- 
http://annevankesteren.nl/
Received on Wednesday, 4 September 2013 20:19:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:37:50 UTC