[Harmony proxies] Opinion on the open issue (How to deal with inconsistent data returned by handler traps?)

David Bruant david.bruant at labri.fr
Fri Mar 18 18:37:41 PDT 2011


Le 19/03/2011 01:11, Brendan Eich a écrit :
> On Mar 15, 2011, at 6:43 PM, David Bruant wrote:
>
>> Unfortunately, currently, on FF4, the property descriptor is
>> rewritten, but with my comments on the code, you can see what results
>> I would expect. But the potential of having my library-specific
>> property descriptor format is here.
>
> Thanks for raising this issue.
>
> The proxies implementation in Firefox 4 is a prototype of a draft
> standard, of course. Between bugs in the spec, future not-quite-bug
> changes to the spec (improvements to go beyond the stated goals of
> proxies), and bugs in our code, we reserve the right to change it at
> any time. Caveat jsdevelopers.
Of course, that's the reason why there is a "non-standard" bannier on
top of the doc
(https://developer.mozilla.org/en/JavaScript/Reference/Global_Objects/Proxy).
Should also be added something saying that the API is subject to change
from version to version?

> This is fine and not a problem, IMHO. Our internal uses for security
> membranes will of course track our changes to proxies.
For the moment, with all the proxy-related discussions, I only see one
thing that could change the way you use proxies and this thing is adding
the proxy as an argument of every trap:
http://wiki.ecmascript.org/doku.php?id=strawman:handler_access_to_proxy
I don't know what is your proxy-based code base size and how proxies are
used exactly, but this sounds like a difficult issue to handle.

> But I wanted to write to assure you that we will make changes that
> TC39 thinks are good for future-proofing, as well as intentional
> changes and of course bug fixes.
Thanks.
In several places I talked about FF4 but it was more to relate my
experience and give concrete examples of spec consequences. Nonetheless,
I knew that the implementation may not currently reflect neither the
spec proposal nor the latest discussions and decisions that could have
happened.

David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mozilla.org/pipermail/es-discuss/attachments/20110319/075314c0/attachment.html>


More information about the es-discuss mailing list