The Big Programming Thread - Page 804

On November 26 2016 05:50 RoomOfMush wrote:

You never declare objects. You declare variables; they are not the same thing as an object. An object is only created when the "new" keyword is used. That is the only way for an object to ever be. (if we ignore black voodoo magic like TheUnsafe and compiler magic like Boxing / Strings)

For example:

String a = new String("Hello World");
String b = a;
String c;
String d = null;
new String("Test");

The above code has 4 variables but only 2 Objects. 2 Variables (a, b) point to the same object. One variable (c) was declared but not initialized; it points to no object at all. This variable is not usable until it is initialized. The last variable (d) is pointing to null which is "no object" in java.
Our second Object is created but has no variable pointing to it. We can not do anything with it anymore because we have no reference to work with.

Its different with member variables (attributes) though. These are ALWAYS initialized because the Compiler enforces that. A member variable of object type is always initialized with null unless you initialize it with something else.

On November 26 2016 05:51 Acrofales wrote:

Was about to answer, but you did it yourself. Not quite sure about the rest. Java changed that shit at some point. It used to be undefined (and effectively initialized to null), and good practice was to initialize them explicitly to null. Then it became explicit for fields (I think in Java 6), and it was unnecessary to initialize fields to null (and you got a warning that that code did nothing... thank you). I have no idea whether variables follow the former or the latter initialization rules. Clearly you are being told to initialize it to null, so it's probably the former. Or you're working with a really old version of Java.

[...] The compiler will assign a reasonable default value for fields of the above types; for local variables, a default value is never assigned. [...]

byte                    0
short                   0
int                     0
long                    0L
float                   0.0f
double                  0.0d
char                    '\u0000'
String (or any object)  null
boolean                 false

On November 26 2016 05:55 travis wrote:


ah.. yeah.. makes sense




okay so it's only local variables that have this problem?

do you know why?? I guess it doesn't *really* matter though it caused me a minor inconvenience, and I don't see the benefit.

On November 26 2016 04:34 RoomOfMush wrote:

No that doesnt make any sense. Binary is a representation of numbers. A computer is a complex physical machine. Everything inside it is currents of electricity. Binary is how we humans decided to interprete it as. Its just a number format that is convenient to use. Binary doesnt happen until you enter the software-level.

On November 26 2016 06:28 phar wrote:

It makes quite a lot of sense. From the hardware perspective, things were quite specifically implemented to give binary signals. When designing a CPU, the people don't put together the pieces and later decide that binary is a good representation. The entire process revolves around being able to isolate a signal as one of two states - binary.

It makes even more sense when we talk about this in the context of the original discussion - data in a computer is most certainly binary under the hood. Your RAM, SSD, or HDD are explicitly storing binary data. Sure it gets a little muddled when you're talking about e.g. tlc flash, which uses 2^3 distinct signal levels to store 3 bits of data... but for all intents and purposes it's binary all the way down, by design.

On November 26 2016 06:44 RoomOfMush wrote:

And I still disagree. Binary is a data representation. It is a way of looking at data but it is not an natural property of the data itself. The representation of the data depends on who is looking at it and for what purpose. A software developer might look at data and see it in terms of objects. A programmer might look at the data and see it in terms of numbers, be it in binary or decimal or whatever. The hardware designer might laugh at this and say the data is all booleans. There is no arithmetics, its all boolean operations on the bit level. The mechanic might see it as voltages or light signals going through a tube or what have you.
What it really is is just data. Its all a whole lot of data. It is binary numbers if you want it to be binary numbers. It is boolean sets if you want it to be boolean sets. But it only is these things as long as there is a human being around to interprete it that way. By its nature its neither.

On November 26 2016 07:35 Blisse wrote:
You're talking about abstractions. The original statement is about the foundations.

Binary is the basic property of everything in a computer system.

We define electric signals by high or low voltage, which we define as 1 or 0. This binary system is the building block of every single thing in a computer. Every abstraction we build is a layer on top of high and low electric signals, a binary system we defined.

On November 26 2016 07:58 BluzMan wrote:


This is a correct but useless definition, the worst of them all.

On November 26 2016 08:04 Blisse wrote:


How is this useless when this is literally how everything works.

On November 26 2016 08:29 BluzMan wrote:

You answered your question with your question. Definitions are useful when they are specific. "How everything works" is the opposite of specific and hence horribly useless.

On November 26 2016 22:15 tofucake wrote:

No, 1 is low voltage and 0 is high voltage. The hardware level is very much binary

On November 27 2016 03:44 Nesserev wrote:

First quality post in a while

EDIT: Anyone got any recommended reads that really blew them away recently?