Pan Fried Programming

(Here's the update -- nothing much is new:
Source Highlight:

Remedial (adj.): intended to rectify, amend, heal.
Panacea (n., myth): goddess of healing, daughter of Aesculapius.
Pansear (n.): Chili's Pokémon.

This remedial lecture will tersely cover a semester's curriculum,
similar to what you have learnt in your high school algebra class,
comprising the fundamentals of programming with synthetic languages
(those that are above machine code).

If you don't know what computer programming is, I would recommend that you study
some tutorials & encyclopedia articles. Much is available on the WWW (Worldwide
Web). The Web is a part of the Internet, and it is the Web you access from your
Web browser when you navigate to a Web page. You could also try one'a them there
"<name of programming language> For Dummies" textbooks: the "For Dummies" books
are excellent "Cliff's Notes"-style crash courses, and each aims to be similar
to a "101" course in the topic advertised.

To make a beginning with any programming language, all you must know is that a
computer computes: your instructions, issued in the program you write, tell the
machine how to progress from its input or initial state to a resultant output or
final state. These instructions look different in different languages -- some
languages require more or fewer -- but every computer program is an algorithm,
or "recipe for computation."

Computers and computer programs can be characterized as finite state automata.
They're like assembly-line robots who know where to weld each sheet of metal.
Also like assembly-line robots, they are "blind" in the sense that they'll kill
you with the soldering iron should you step between it and the sheet.
Computing machines do what they're told, even when it is particularly stupid:
that's why computer viruses, worms, and computer espionage exist.

In simplest terms, the computer's processing units receive some numbers and an
instruction that says what mathematical operation to execute, then operates:
like a calculator. High-level programming languages are more synthetic, like a
human language is, and comprise such ideas as objects (amalgamations of data) &
functions (modular sub-routines). Compilers or interpreters read these languages
and translate them into machine instructions, simplifying the lengthy series of
instructions necessary to make the calculator execute these difficult tasks.

In a high-level language, there are few technical concerns.
You can begin immediately with the abstract concepts.
Here are some:

As in algebra, a variable is a name that represents a value.
As in solving a system of equations, values are typically assigned to some vars
and the value of the other variables is computed using the values given.
For example, in Javascript:
    var a = 2;
    var b = a + 2;
The variable <b> is now equal to 2 + 2. Similar operations function similarly.
In Javascript and other very-high-level languages, variables aren't only scalars
and can point at any object. They're like placeholders for procedure.
Although "variable" implies a value stored in memory, and "identifier" only its
mnemonic, the words "variable" & "identifier" used loosely mean about the same.
    "Just don't try that with the Captain."
        -- Geordi LaForge, to Data, _Star Trek: the Next Generation._

These are important ideas that are abstracted away in VHLLs. A pointer stores an
address in memory, for a later indirect read/write or similar operation. In HLLs
a pointer/reference accesses an object directly instead of copying its value.
You'll rarely have to make the distinction in Javascript; but, for example:
    var a = new Array(1, 2, 3); // a[0] == 1, a[1] == 2, a[2] == 3
    var b = a; // Incidentally, b === a, and that is why in the next line...
    b[0] = 555; // ... b[0] == 555, and now a[0] also == 555!
As opposed to:
    var c = new Array(3); // c is a new array of length 3
    c[0] = b[0]; c[1] = b[1]; c[2] = b[2]; // copy scalar values one-by-one
    c[0] = 0; // c[0] == 0, but b[0] remains == a[0], which remains == 555.
    var d = 2;
    var e = d;
    e = 4; // e == 4, but d is still == 2.
As you can see, operating on an object (such as via array subscript operation)
changes the object, even if the object is pointed by multiple variables.
Likewise, objects passed as the argument of a function are passed by reference:
they aren't simply copied, and operating on the argument within the function is
equivalent to changing the object, whose scope is above that of the function.
Some high-level languages, like C, permit you to explicitly specify what is a
pointer or reference, which eliminates some of this confusion but requires more
exacting attention to detail in your design specification.

The state of a program is the value of all its variables, the current location
within the instruction set, and the environment of the operating system (or the
interpreter). In Javascript, within Web browsers, the browser typically provides
access to some of its state via the Document Object Model.

Heuristics, or "guesswork," could not exist if there were no way to execute some
different code depending on the state of the program. Furthermore there are some
mathematics you can't write as exactly one set of instructions that produces one
semantic value: for instance, a function defined only on an interval, or an even
root of a positive number. In this circumstance, you are writing branches:
    if (5 > 10) { /* of course, the code in this block never happens. */ }
    else if (2 < 0) { /* neither does this, b/c cond'n is also false. */ }
    else { /* but all of this code happens, because the others didn't. */ }
... One of the branches executes, and the others don't.
The part in parentheses is the "conditional statement:" it's evaluated as either
"true" or "false," like in Boolean logic. 

Identifiers are only valid within the block (curly brackets, or { ... }) where
they were declared. Well, they're supposed to, anyway. Therefore, if you declare
a variable inside a function, you can't use it outside of the function or within
another function. Why would you want to, anyway? The next time you invoked the
function, the value of the variables you were using in there would change again.

Computers are great at repetition. Loops repeat a set of instructions: they are
typically written as a prefix, conditional, postfix, and body. For example:
    for (var T = 10; T > 0; T--) { alert("T minus " + T); }
... which counts down from ten to one with annoying alert popups.
While or do-while loops have only conditions & bodies.
A loop is an example of an "iterative algorithm." Each time the loop's body is
executed, it's called an "iteration." In computing fractal geometric patterns,
"iteration" means more like "recursion:" which, see below.

A function is a modular segment of your program: a sequence of computation that
is repeated a few times, or can be reused as part of another computation.
Functions are "invoked," or called by name, with values supplied as arguments,
and return a value, similarly to how functions behave in algebra. When declaring
a function, you'd typically write the name of the function followed by its args
in parentheses and then the function body. For example, again in Javascript:
    function intp (N) { return (N % 1) == 0; } // integer predicate
... which returns true if N is probably an integer, or whole number:
    if (intp(5)) { alert("Yes. 5 is probably an integer."); }
    if (intp(5.55)) { alert("This box never appears..."); }
    else { alert("... but this one does, because 5.55 is a floater."); }
(Floating-point numbers are inaccurate, in Javascript as likewise elsewhere.)

A function that invokes itself is a recursive function. Any function invoking an
other function, which subsequently causes the original function to be invoked
again, causes a recursion-like situation that I think is called "re-entrancy."
It is essential to note that _any_ and _every_ recursive function you can write
for a computer to execute can be rewritten as an iterative algorithm. The proof
of this is complex: it follows from Alan Turing's model of finite state automata
and the read-execute model of arithmetic and logic units (CPUs), and basically
asserts that you'd never be able to execute recursion if you couldn't do it by
reading one instruction at a time. In other words, each time your function calls
itself again, it's simply storing its state in memory temporarily while the
machine executes a fresh copy: after the copy is done, the former state is re-
loaded, and execution proceeds from there. This is achieved with stacks: data
structures that grow in size as more is "pushed" onto them, and shrink when some
is "popped" off of the top.

An object is a collection of data that comprises a several datum. That is, when
data are related to one another, they can be envisioned as a "shape" or "motion"
that is the sum of its parts. For example, a vector has direction and magnitude;
an individual has a first and last name; a parser has an input stream, a stack,
and a procedure. In Javascript, you'd write something like this:
    function Blah (newz) { if (newz) { this.z = newz; } return this; }
    Blah.prototype = new Object();
    Blah.prototype.z = 555;
    Blah.prototype.tell_me_z = function () { return this.z; }
    var a = new Blah(222), b = new Blah(); // a.z == 222; b.z = 555.
... syntax varies among languages. Essentially an object is a data structure
containing some members ("variables" attached to the object, like Blah::z above)
and, if the object is a class, some methods (functions, like ::tell_me_z).

One thought on “Pan Fried Programming

Be careful what you say. The U.S.A. is not as free a country as you have been led to believe.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s