apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Greg Marr <gr...@alum.wpi.edu>
Subject Re: Thread pool prototype
Date Wed, 03 May 2006 14:10:27 GMT
At 10:00 AM 5/3/2006, Joe Orton wrote:
>On Wed, May 03, 2006 at 09:41:14AM -0400, Greg Marr wrote:
> > At 05:32 AM 5/3/2006, Joe Orton wrote:
> > >> +    me->cnt_max = max_threads;
> > >> +    me->idle_max = init_threads;
> > >> +    rv = apr_thread_mutex_create(&me->lock,
> > >APR_THREAD_MUTEX_NESTED,
> > >> +                                 me->pool);
> > >> +    if (APR_SUCCESS != rv) {
> > >
> > >Personally I find the "constant != variable" style very 
> distracting,
> > >not sure if this is a widely held view...
> >
> > FYI, I use it almost exclusively, because if you mean to type
> >
> > if (APR_SUCCESS == rv)
> >
> > and instead you type
> >
> > if (APR_SUCCESS = rv)
> >
> > the compiler will catch it for you.  If you normally do variable 
> ==
> > constant instead, you have no such protection.
>
>GCC has issued a warning for "if (foo = 2)" forever, if you aren't 
>running your code through gcc -Wall regularly you have more serious 
>problems anyway, certainly that's not worth sacrificing readability 
>for IMO.

I mostly work on Windows, and VC++ only recently added a warning for 
that.  I've been using this style for so long that it doesn't 
sacrifice readability, and in fact, the var == const form looks wrong 
now.  However, this is entirely a personal preference.


Mime
View raw message