On Thu, Oct 11, 2012 at 11:37 PM, Guglielmo Braguglia < guglie...@braguglia.ch> wrote:
> Hi Pete, > > if you declare a variable "global",the variable IS GLOBAL FOR ANY STACK > and is "*persistent*". This means that the variable and its value exist > regardless of the stack that created. If you liketo see do just a simple > test ... > Wow, so a global declared in one stack can be accessed by a completely different stack, even if the declaring stack is not in memory. That seems like a recipe for potential disaster, especially if you are writing plugins that are used by customers to access their own stacks. > ... create a stack and create different scripts (e.g. some scriptfor some > buttons). On the different scripts define "local" variable on the TOP of > the script, before handlers. Assigns them a value. Now, if you insert some > breakpoints (on the different scripts) you will see, still looking into the > "Variables" TAB, that, when the script stop because reach the breakpoint > (so the script is the current script), you will find the local variable of > the script with the correct value. > I use script local variables a lot and often instead of global variables. A script local variable plus a set of handlers to read and write from it seems a lot easier to deal with than global definitions in every script that needs them, especially now I know about the cross-stack global availability issue mentioned above. I've also taken to using script local variables/access handlers instead of using custom properties in locked stacks so users can't change the values of critical information by simply bringing up the Inspector window. I still find it odd that password protecting a stack prevents access to its scripts but not its custom properties; there at least ought to be an option to apply password protection to custom properties. > > Please note that, also if this rules remain valid, the behavior of > LiveCode is different if your "Strict Compile Mode" is ON or OFF !!! ... > and this creates more confusion !!! > > With "Strict Compile Mode" ON ... you are obliged to declare the variable > and LiveCode verify the "collisions" and give you error if one local have > the same name of a global, BUT, if "Strict Compile Mode" is OFF, LiveCode > don't verify the "collisions" and the same variable can be a local or a > global depending on whether you declare it "global" or not :o ... and this > is really ridiculous !!! > That's the crux of the question I posted in response to Mark's post. The whole implementation of globals in LC is really strange. Especially since it appears that LC "knows" about a global as soon as it is declared - why then force people to re-declare them in every script that addresses them. I rarely use globals now but "rarely use" will change to "never" pretty quickly! > > Hope this clarify the situation ... :-) > Yes, thanks Guglielmo. > > Guglielmo > _______________________________________________ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-livecode
