Addition of a global namespace function?

Alex Russell alex at dojotoolkit.org
Fri Dec 4 14:20:14 PST 2009


On Dec 4, 2009, at 2:13 PM, Brendan Eich wrote:

> On Dec 4, 2009, at 2:11 PM, Alex Russell wrote:
>
>> me:  the problem is that people can still omit var
>> you: everyone will run it through a verifier
>> me: programmers are still human, particularly web developers. They  
>> don't run verifiers. In general we still have the problem.0
>> you: everyone will run it through a verifier
>> me:  really? And if they're going to accept the constraint anyway,  
>> wouldn't it be better if we just created a construct where the  
>> default is changed?
>> you: but everyone can just run it through a verifier!
>> me:  um...
>>
>> Lawls.
>
> Let's back up.
>
> You and I know we can't incompatibly change the language. So "use  
> strict" or better ("use lexical scope", or just the default in a  
> future opt-in version based on the Harmony proposals) will be  
> necessary to get the right non-default-but-built-into-browsers  
> behavior, where free variables are static errors. Anything else is a  
> new Web, e.g. Flash or Silverlight.
>
> Unless you have a better idea?

Nope. I'm only suggesting that we need to get to "use better_thing"  
SRTL and that ES5 strict mode doesn't fix the problem in an  
appreciable way for most content. I expect we're in agreement.

Regards


More information about the es-discuss mailing list