realbasic-betas.mbox
[Top] [All Lists]

NewClassInstance related to RBScript potential bug?

To: REALbasic Betas <realbasic-betas at lists dot realsoftware dot com>
Subject: NewClassInstance related to RBScript potential bug?
From: Theo <delete at softhome dot net>
Date: Wed, 27 Mar 02 00:35:26 +0100
> Ah - wait. RS can't fix this easily because it does not know which
> classes will be used. Darn! And with the planned optimizing linker
> might become even worse.

Well, even if RS implemented this feature, they would have the same
problem as a plugin developer did. Lets say we made a new class, from a
string read out of a file in the hard disk. Now there is no way RS
could know what to strip out or not in this case. In this case, it
would be up to the program author to decide which classes always get
compiled and which don't.

This basically means we need some way to inform RB of the classes we
want compiled or not.

I agree that we'd need to see if users need it. It was just an off the
wall "neat" idea, but we should remember that the most limited resource
a programmer has is time and energy. Off the wall ideas don't require
any energy unless implemented. So.... I guess I'll just keep this idea
in mind.

There is always a work around in case you wanted to do it yourself.
Simply make a module called "LinkModule", with a method named
"LinkMethod", and from that method acess functions of the classes that
you want linked.

Also, finally, I can see that we DO need this feature in RB, already.
As you said, we can do this in RBScript. So therefor, the new linker
will need some way to inform RB of the classes to link!

RS, I think we pre-emptively found a bug you might have missed in
RBScript that can only be fixed with some way of informing RB :o)

--
    This email was probably cleaned with Email Cleaner, by:
    Theodore H. Smith - Macintosh Consultant / Contractor.
    My website: <www.elfdata.com/>



<Prev in Thread] Current Thread [Next in Thread>
  • NewClassInstance related to RBScript potential bug?, Theo <=