[Stk] noob pointer question about the tutorials

[malik martin]

oops, this never seems to go back to the group, just in case anyone
else is a noob :)

"hehe, thanks for the detailed explaination!
i'm used to a referene being declared like "&reference" the many
different ways to declare a pointer or refference can be really
confusing to me, so yeah, i do need to read up on that and their
importance.

and and what is "being cricket?" hehe :)

also, haven't read much about exception handling to be honest. will
definately give that a look. i could probably use it big time!
spent 2 days coding a vst and cant firgure out why it's crashing a
host :P and didn't stop to compile to see if everything was going
well. :/

thanks again :)
Malik"

> On Nov 14, 2007 5:07 AM, Richard Dobson <richarddobson@blueyonder.co.uk> wrote:
> >
> > malik martin wrote:
> > > why in "hello Sine!' was FileWvOut duplicated as a class instance
> > > named "output" and in realtime auido blocking, an instance was
> > > created as a pointer, "*dac"? i read that the class is protected, and
> > > i've read about pointers and classes a bit. not that i can truly say
> > > i remember it but, why so much protection? why not make an instance?
> > > will this instance affect the base class or something?
> > >
> > > oh and what's this address operator doing?? "( StkError & ) " the
> > > address of what? it's by itself. and for that matter i never really
> > > understude the return codes 0, 1, etc. i was just taught to know
> > > that, typically, return 0; exits the program. the reason i as is
> > > cause of  " exit(0);"
> > >
> >
> >
> > If the constructor of a class can throw an exception (which wvOut can)
> > it is best (necessary even) to create it via new inside a try block, so
> > the exception can be caught.  Where (as in most cases) the constructor
> > does not throw an exception (the default constructor of fileWvOut will
> > always complete), you can declare an instance on the stack, as "hello
> > Sine" does. It is a matter for debate whether a constructor that can
> > throw an exception (especially the one and only default constructor) is
> > a good thing or not.
> >
> > You need to read up on exceptions in C++ - much of it will
> > not make sense until that aspect is understood.
> >
> >
> > The notation "StkError &" signifies a C++ reference variable; which you
> > also need to read up on in great detail (used for all sort of things,
> > but very much in exception handling). It is a ~sort of~ pointer that is
> > ~guaranteed~ to point to an valid object (enforced by the language);
> > whereas a normal C pointer can be declared but not initialised, so could
> > be NULL or point to rubbish memory. You ~never~ have to test a reference
> > argument or variable to see if it is valid or not. It is scarcely
> > possible to write or read non-trivial C++ without knowing about references.
> >
> > The return value of a program is often irrelevant, but it can be read by
> > the parent process (which might be a unix shell script, among other
> > things), which can read the return value to decide the next action.  By
> > convention, a return value of 0 signifies success, while any other value
> > (positive or negative)  signifies some error.  Thus at the end of a
> > program, it is usual to write "return 0" to signify successful
> > completion of the process.
> >
> > "return" has no special meaning in terms of exiting a program; you use
> > it in any function that is defined to return a value. main() itself is
> > defined by C and C++ to return an int, so you use return at top level to
> > terminate the program. "exit" is a shortcut (some would say a lazy one),
> > to enable a program to terminate from some arbitrary point deep within it.
> >
> > It has to be said that wrriting exit(0) inside an exception handler is
> > not really cricket (speaking as an Englishman!) - it should be exit(1)
> > or something, to communicate the error to the outside world (the
> > argument to exit() becomes the return value of the program). But of
> > course programmers are notorious for bending such conventions to their
> > personal will!
> >
> >
> >
> >
> > Richard Dobson
> >
> >
> >
> > _______________________________________________
> > Stk mailing list
> > Stk@ccrma.stanford.edu
> > http://ccrma-mail.stanford.edu/mailman/listinfo/stk
> >
>