Defining Pet Features and Essentials

[Ender Nafi Elekcioglu]

Geoff:
> I just tested python on an ipad and it took about 30 seconds for fib(33).

30 seconds for Python, you say;
my, my, the plot’s getting thicker and thicker.

I tested with Livecode just now for an up-to-date comparison.
It’s 11 seconds for that base algorithm on iPhone 5S;
and 437 *micro* seconds for your optimized algorithm.

Btw, it’s 103 seconds and 3 milliseconds, respectively, on the iPod 4 which is a very weak device, of course.

iPhone 5S is probably more powerful than iPad -depending on its model-
but even if that is the case, I don’t think it’ll be much worse.

So, in my opinion, it won't be unjust to say that Livecode beats Python on mobile fair and square.
That’s kinda relief for me, I mean, it increases confidence to my choice of platform.


Geoff:
> As long as LC is dynamic (not compiled) it is unlikely to be as fast as C.

Expecting Livecode or any other RAD tool to be as fast as any low-level language, especially C, is a dream, of course; I’m aware of that.
Even Objective-C can’t be compared to C; I’ve read many showcases where Objective-C developers fall back to C for select demanding tasks.

However, your statement of *dynamic vs compiled* caught my attention.

I have no formal education in computer sciences, whatsoever; so I don’t know the core difference(s) between those concepts.

If I should understand by “compiled” 
that Livecode won’t be “live” anymore 
and that I have to wait my script's compilation to see my work;
I’d happily sacrifice it for any level of performance bump.

But if “dynamic and not-compiled” brings us 
the convenience of flexible type variables, ease of syntax, scripting the objects individually, etc.;
well, that’s a different story; I should shut my mouth immediately, then :)


~ Ender