[Openmcl-devel] process-run-function and "dynamic vs lexical"

[Taoufik Dachraoui]

On Oct 22, 2009, at 6:08 PM, Greg Pfeil wrote:

>
> On 22 Oct 2009, at 11:42, Taoufik Dachraoui wrote:
>
>> Just for completeness I include the example:
>>
>> ? (defvar y 1)
>> Y
>> ? (let ((x 2) (y 3))
>>   (labels ((f () (format t "~%x=~A y=~A~%" x y)))
>>     (ccl:process-run-function "test" #'f)))
>> #<PROCESS test(3) [Reset] #x8D43546>
>> ?
>> x=2 y=1
>> ?
>>
>> let creates new variable binding for the names x and y and these
>> bindings are lexical unless they are declared special.
> ...
>> Where is my mistake, please note here that I would like to reconcile
>> my understanding with what I see while running ccl code.
>
> If you check Ron's guide to specials (http://www.flownet.com/ron/specials.pdf 
> ), the section "The pervasiveness of DEFVAR" explains what you're  
> seeing.
>
Does this pervasiveness is a fatality?

Looking at CLHS the only thing I could see so far is that (DEFVAR name  
init)
establish  name as a dynamic variable. I could not find anything about  
the
pervasiveness of DEFVAR.

Also, as explained in my previous post, the let block create new  
lexical bindings
and since the y name was not declared special I  do not see why the  
DEFVAR
has to complicate things.

Please look again, and pay attention to the name substitutions and try  
to tell me
why my reasoning is wrong.

Kind regards
Taoufik

> "Because DEFVAR does more than just establish a dynamic binding for  
> X. It pervasively declares all references to X and all subsequent  
> bindings for X to be dynamic (or special -- same thing). In other  
> words, DEFVAR turns its argument (permanently and pervasively) into  
> a special variable."
>
> So the Y <= 3 binding is actually for the special variable, but the  
> binding of the 3 is thread-local, so your function run in its own  
> thread still sees the value 1.
>