Re: [Chicken-hackers] [PATCH] disable warning for let-bound keyword

[Felix]

From: Christian Kellermann <address@hidden>
Subject: Re: [Chicken-hackers] [PATCH] disable warning for let-bound keyword
Date: Wed, 30 Nov 2011 14:32:46 +0100

> * felix winkelmann <address@hidden> [111130 13:49]:
>> From: Christian Kellermann <address@hidden>
>> Subject: Re: [Chicken-hackers] [PATCH] disable warning for let-bound keyword
>> Date: Wed, 30 Nov 2011 11:12:03 +0100
>> 
>> > * Christian Kellermann <address@hidden> [111130 11:00]:
>> >> No crash with this, thanks. There is a similar call in compiler.scm
>> >> around line 609, is it safe to call there?  If it is I will push
>> >> out the patch with the above change.
>> > 
>> > I have pushed the corrected patch. If the compiler.scm hunk is an
>> > issue we will make another one..
>> 
>> The compiler has access to srfi-1 (it is loaded in this case).
>> But I will change it to use the simpler code as well.
> 
> Ah ok, is it due to known arity that (lambda (x) (car x)) is more
> efficient than (map car ...) ?

Forcing inlining is only to guard against redefinition of "car"
(this may be an issue in the interpreter, particularly for new
users who like to experiment). But actually both forms should
expand into fully inlined code, since "map" is a compiler-macro.
I haven't checked, I'm just careful.


cheers,
felix