Pass variable by reference

[Rustom Mody]

On Wednesday, May 7, 2014 5:16:16 AM UTC+5:30, Chris Angelico wrote:
> On Wed, May 7, 2014 at 7:00 AM, Mark H Harris wrote:

> Is this code mutating or rebinding?

> x = 1.1
> y = 2.2
> x = x + y


Heh! Neat example!


> What language did I write that in? Is there really a fundamental
> difference between languages in which that is equally valid syntax and
> does exactly the same thing? Apart from the fact that BASIC defaults
> to single-precision float (in the absence of a hash suffix), Python
> uses double-precision, and REXX uses decimal strings, this sequence
> will work identically in all of them. Does BASIC have variables?
> Presumably, since you can pass them by reference. Does REXX? You can't
> pass by reference (all you can do is EXPOSE, which is more like
> scoping rules, only it isn't); a C function can fiddle with any named
> variable in its caller's scope, so a common way to "return" multiple
> values is to pass the name of a stem (which isn't an array, and isn't
> a dict/mapping, and isn't really an object at all) into which the
> results will be placed. Python has objects and names, and changes
> which object a name is bound to.

> The mechanics differ, but fundamentally, x started out as 1.1 and
> ended up as 3.3. That means x varied in value, and is thus a variable.
> The differences are relatively insignificant.

Wrong conclusion!

These 3 lines look the same and amount to much the same in python and C.

But as the example widens to something beyond 3 lines, the difference
will become more and more significant

OTOH...

As for meaning of the word 'variable,' the way mathematicians use it
seems to date to the 16th century:
http://en.wikipedia.org/wiki/Variable_%28mathematics%29#Genesis_and_evolution_of_the_concept

That is, its closer to the binding notion of python than the imperative notion
of C like languages.

However if we go further back, the book by Al-Khwarizmi is about
al-Jabr
ie the name of the guy who invented algebra is used for 'algorithm'!!

I find this ironic, given this current discussion:
- algorithm corresponds to the mutating notion
- algebra corresponds to the non-mutating notion