On Fri, 28 Nov 2008 17:14:55 -0700, George wrote:
> By trade, I'm a carpenter, and I've been implementing a greener heating
> solution for the house I'm flipping including fans that are to be
> controlled by an integrated circuit. I expect this solution to help the
> house, help the planet, and put more green in my wallet.
>
> It is in the programming for, say, a thermistor, that volatile comes into
> play. I thought of doing the whole project in C, which, I thought, was the
> best tool for this type of thing. I've now stumbled upon an opinion that
> volatile in C99 is not a good thing, and now I have a couple questions.
>
> q1) Is there anybody around with a serious criticism of volatile in C99?
>
> q2) What changed about volatile between C90 and C99?
>
> I have 1256.pdf on my machine now. Thanks for your comment.
%- >I've been looking at volatile from the C side, and I like the fortran
%- >model, as I understand it in Adams. This is clearly an issue where
fortran
%- >best follows C99.
%-
%- No, it isn't. Fortran VOLATILE is less disgustingly undefined than
%- C volatile, and following that is insane. The latter is currently
%- being touted for use in combination with parallelism (POSIX threads
%- etc.), and the combination simply doesn't work. If you are VERY
%- lucky, your implementation will have extra documentation that says
%- what it actually does, but I have never seen that.
%-
%- I put in a couple of "interpretation requests" on Fortran VOLATILE,
%- to which the standard's editor's response was to make a grimace.
%- They are FOUL. Please ask if you want me to post them.
Would anyone object to the custom quoted text?
--
George
The course of this conflict is not known, yet its outcome is certain.
Freedom and fear, justice and cruelty, have always been at war, and we know
that God is not neutral between them.
George W. Bush
Picture of the Day
http://apod.nasa.gov/apod/