[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: bindings and assignments (was: Re: continuations)




On Thursday, August 14, 2003, at 08:53  AM, John Clements wrote:

>
> On Wednesday, August 13, 2003, at 10:53  PM, Shriram Krishnamurthi 
> wrote:
>
>> Vadim Nasardinov wrote:
>>
>>> If you don't use the "final" keyword, then the variable can be
>>> reassigned (in the set! sense).
>>>
>>> Does that qualify as a syntactic distinction?
>>
>> I'd forgotten about this use of "final" (maybe because I don't find
>> the pun quite so evident).  Yes, that would qualify as a syntactic
>> distinction.
>
> I respectfully disagree.  Insofar as the 'final' keyword may also be 
> applied to an object's member variables to prevent them from being 
> mutated, I believe that it _increases_ the mental conflation of 
> bindings and fields.
>
> I am disappointed but not surprised as I read the (PLDI 200) C-- 
> semantics to see that even at an abstract syntax level, Ramsey et. al. 
> choose to use a single form for set! and set-box!.  Of course, this 
> means that he must have two different reduction rules for the same 
> abstract syntax form... one if the assignment is to a variable, one if 
> it's to a memory location.  Bleah!
>
> john

Forgive me; I'm beating the binding vs. store dead horse here, rather 
than the binding vs. binding-mutation dead horse.  Wrong mailing list, 
wrong dead horse.

john