Garrett D'Amore writes:
> I'm not sure if the Style Guide makes that difference. I don't know.
> Frankly, this is one of my least favorite rules in the guide, because I
> happen to think "if (!foo)" is clearer than "if (foo != 0)". But I
> didn't write the guide, or get to vote on it -- but I do assume that I'm
> bound to follow it.
>
> If I've misinterpreted what it says, I'd be happy to be corrected,
> because to *my* eyes, I agree with Peter.
I think there's general agreement that treating a bitwise flag test as
a boolean results in nicely readable code, and should be encouraged.
In fact, I'd go so far as to say that comparing those expressions
against zero needs to be outlawed. "== 0" is very hard to read as
"not set," while "!= 0" has a hint of not-i-ness about it.
(Whether we ought to have bitwise values at all or should be using
explicit booleans is probably a separate matter, and doesn't matter
much given the context here.)
What the Style Guide is trying to prevent is the use of weird
non-boolean values as though they were boolean. This is fairly ugly,
because it just doesn't read right out loud:
if (foo = malloc(10)) {
/* looks like I can allocate 10 more bytes! */
}
If foo is ... what? This is much worse:
if (open("/dev/null", O_RDWR)) {
/* now why is stdin already open? */
}
The code needs to be readable.
--
James Carlson, Solaris Networking <[EMAIL PROTECTED]>
Sun Microsystems / 35 Network Drive 71.232W Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677
_______________________________________________
driver-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/driver-discuss
