- From: David Bruant <bruant.d@gmail.com>
- Date: Tue, 06 Nov 2012 10:51:51 +0100
- To: Boris Zbarsky <bzbarsky@MIT.EDU>
- CC: Cameron McCormack <cam@mcc.id.au>, Alex Russell <slightlyoff@google.com>, public-web-security@w3.org
Le 05/11/2012 23:35, Boris Zbarsky a écrit : > On 11/5/12 2:16 PM, Cameron McCormack wrote: >> If it's the latter, then I think either Boris' suggestion of having two >> interfaces, and stating in prose which one is implemented on Document, >> would be OK. I think it'd be fine too to just make it not readonly in >> the IDL and for there to be a prose hook to do all the right >> throwing-or-ignoring things that you would normally get from assigning >> to an accessor property without a setter. > > I don't see how the latter can really work. As a simple example, if I > grab a setter out of a property descriptor and then later call it, > which code's strict mode is relevant? The code that grabbed the > setter or the code that called it? Is there anything that requires an > ES impl to actually keep track of strict mode for function calls? > Because I don't think Spidermonkey does anything like that.... Strict mode is a static thing. If a function is defined (and instantiated) in a strict portion of a program (or is a strict function), then, regardless of who's grabbing and who's calling, the function applies strict rules. I think the essence of strictness should be kept. David
Received on Tuesday, 6 November 2012 09:52:35 UTC