[MPlayer-cvslog] r25777 - in trunk/libswscale: rgb2rgb_template.c swscale-example.c swscale.c swscale_altivec_template.c swscale_template.c yuv2rgb_altivec.c

[Benoit Fouet]

Hi,

Uoti Urpala wrote:
> On Thu, 2008-01-17 at 11:56 +0100, Benoit Fouet wrote:
>   
>> Reimar Döffinger wrote:
>>     
>>> On Thu, Jan 17, 2008 at 10:26:34AM +0100, Benoit Fouet wrote:
>>>       
>>>> I guess that removing parentheses is ok only if it doesn't create new
>>>> warnings, or ?
>>>>         
>
>   
>>> Actually, my and AFAIK a few other's opinion is that the warning should
>>> be disregarded in the && vs. || case, since the evaluation order is
>>> obvious to anyone who ever had to do with any kind of logic stuff (or
>>> IOW it is generally known, accepted and consistent).
>>>       
>
>   
>> there are some of if ((a || b) && (c || d)) in the code, but I won't
>> change them to a version with less parentheses.
>>     
>
> "Disregarding warnings in the && vs || case" means not using parentheses
> for "a && b || c && d". Rewriting "(a || b) && (c || d)", which does
> require parentheses when written with those logical operators, to a form
> like "a && c || a && d || b && c || b && d" that does not require
> parentheses is something else entirely.
>
>   

indeed, this was what I meant !
thanks for correcting me

BTW, I'm not sure the && precedence on || is something that is true in
all logic stuff: for instance, IIRC, in shell, they have the same priority

-- 
Ben
Purple Labs S.A.
www.purplelabs.com