unicorn-utterances/content/blog/non-decimal-numbers-in-tech/index.md

title, description, published, authors, tags, attached, license

title	description	published	authors	tags	attached	license
How Binary and Hexadecimal Work: An introduction to non-decimal number systems	Learn how to convert decimal to binary and hexadecimal, how CSS colors are calculated, and how your computer interprets letters into binary	2019-11-07T05:12:03.284Z	crutchcorn	binary hexadecimal		cc-by-nc-sa-4

Computers - on a very low level - are built upon binary (ones and zeros). Think about that - all of the text you're reading on your screen started life as either a one or a zero in some form. That's incredible! How can it turn something so simple into a sprawling sheet of characters that you can read on your device? Let's find out together!

Decimal

When you or I count, we typically use 10 numbers in some variation of combination to do so: 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9.

When you count to 10, you're really using a combination of 1 and 0 in order to construct a larger number that we cognizantly recognize. The number 10 persists in our minds even when we have it written out; ten.

Knowing that we can separate the number from our thoughts allows us to categorize the number in a further manor, breaking it down into smaller groupings mentally. For example, if we take the number 34, for example, we can break it down into three groups: the ones, the tens, and the hundreds.

For the number 34, we break it down into: 0 hundreds, 3 tens, and 4 ones. We can then multiply the higher number with the lower number (the column they're on) to get the numbers 30 (3 tens) and 4 (4 ones). Finally, we add the sum of them together to make the number we all know and love: 34.

This breakdown showcases a limitation with having 10 symbols to represent numbers; with only a single column, the highest number we can represent is 9. Remember that the number 10 is a combination of 0 and 1? That's due to this limitation. Likewise - with two columns - the highest number we can represent is 99

Binary

Now this may seem rather simplistic, but it's important to demonstrate this to understand binary. Our numerical system is known as the base 10 system. Called such because there are 10 symbols used to construct all other numbers (once again, that's: 0, 1, 2, 3, 4, 5, 6, 7, 8, and 9). Binary, on the other hand, is base two. This means that there are only two symbols that exist in this numerical system.

For the latin enthusiasts, binary comes from "binarius" meaning "two together". Deca, meaning 10, is where "decimal" comes from

Instead of using numbers, which can get very confusing very quickly while learning for the first time, let's use **X**s and **O**s as our two symbols for our first few examples. An X represents if a number is present and we should add it to the final sum, an O means that the number is not present and we should not add it. Take the following example:

In this example, both 1 and 2 are present, so we add them together to make 3. You'll see that since we can only have a value present or not present — because we only have two symbols in binary — this conversion has less steps than using decimal. For example, if you only wanted the number two, you could simply mark the 1 as "not present" using the O

You can even replace the two symbols with 1 and 0 to get the actual binary number of 10 in order to represent 2

So how does this play out when trying to represent the number 50 in binary?

As you can see, we create columns that are exponents of the number 2 for similar reasons as exponents of 10; You can't represent 4, 8, 16, or 32 without creating a new column otherwise.

Remember, in this system a number can only be present or not, there is no 2. Because of this, without the 4 column, there can only be a 1 and a 2, which makes up 3. Continuing on with this pattern: without an 8 column, you can only have a 4, 2, and 1 which would yeild you 7.

Once each of these exponents is laid out, we can start adding 1s where we have the minimum amount of value. EG: Is 64 less than or equal to 50? No. That's a 0 Is 32 <= 50? Yes, therefore that's a 1 50 - 32 = 18 Moving down the list, is 16 <= 18? Yes, that's a 1 18 - 16 = 2 Is 8 <= 2? No, that's a 0 4 <= 2? No, that's a 0 as well 2 <= 2? Yes, that's a 1 2 - 2 is 0. That means every number afterwards (in this case only 1 is left) is not present, therefore is a 0.

Add up all those numbers:

Column	Value
64	0
32	1
16	1
8	0
4	0
2	1
1	0

And voilà, you have the binary representation of 50: 0110010

Author's note:

While there are plenty of ways to find the binary representation of a decimal number, this example uses a "greedy" alogrithm. I find this algorithm to flow the best with learning of the binary number system, but it's not the only way (or even the best way, oftentimes).

Hexadecimal

But binary isn't the only non-decimal system. You're able to reflect any numerical base so long as you have the correct number of symbols for that system. Let's look at another example of a non-decimal system: Hexadecimal.

Hexadecimal is the base 16 number system.

Hexa meaning "six" in Latin, deca meaning "ten", combining to mean "sixteen".

Now you may wonder how you can count to 16 in a single column when we only use 10 numbers. The answer, to many developers, is to fill the remaining last 6 with other symbols: letters.

0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F

These are the symbols that makeup the hexadecimal numerical systems for many devs. A, in this case, represents the number 10. F being the number 15. In this numerical system, there are the sixteens, the two-hundred fifty sixs (gathered by multiplying 16 by itself — 16 ^ 2), and other exponents of 16.

Given this information, how would we represent the number 50?

Assuming we have a ones column, a sixteens column, and a two-hundred fifty sixths, let's do a similar method of calulating the number as we did with binary.

Is 256 less than or equal to 50? No. That's a 0 Is 16 <= 50? Yes. So we know it's at least 1.

Now, how many times can you put 16 in 50?

16 * 2 = 32 and 32 <= 50, so it's at least 2

16 * 3 = 48 and 48 <= 50 so it's at least 3

16 * 4 = 64. However, 64 > 50, therefor the sixteenth place cannot be 4, therefore it must be 3

So now that we know the most we can have in the sixteenth place, we can subtract the sum (48) from our result (50)

50 - 48 = 2

Now onto the ones place:

How many ones can fit into 2?

1 * 1 = 1 and 1 <= 2, so it's at least 1

1 * 2 = 2 and 2 <= 2 and because these number equal, we know that there must be 2 twos.

Now if we add up these numbers:

Column	Value
256	0
16	3
1	2

Why 256?

While reading through this, you may wonder "Where did the 256 come from?". Let's take a step back to anaylize this question.

If you recall, we use the 15 symbols in hexadecimal:

0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F

If we reflect these numbers in a single digit (1 number column), the biggest number we can reflect is F: 15.

This is similar to how the biggest number we can reflect with a single digit in the decimal system is 9.

In order to add a number larger than 15 in the hexadecimal system, we need to add another decimal/column. This column would represent the sixteens place. Having F in this column, the highest we can represent, and an F in the ones would allow us to have FF: 255. As a result, we need to add a two-hundred-fifty-six column to represent any numbers higher.

For those that have experience in algebra, you'll notice that these are all exponents of each other.

Just as 100 is 10^2 for the decimal system, 256 is 16^2. We can follow this pattern to the next number in the hexadecimal column: 4096, which is 16^3. You can even apply it to 1 which is 16^0

Binary works in the same manor. The first 5 columns/digits of binary are: 1, 2, 4, 8, 16 . These numbers align respectively to their binary exponents: 2^0, 2^1, 2^2, 2^3, 2^4

It's also worth noting that decimal numbers can be written out the same way.

732 for example, in base 10, can be written as 7x10^2+ 3x10^1 + 2x10^0

To Binary

Remember that at the end of the day a hexadecimal number only reflects a number underneith. Just as we're able to convert from binary to decimal, we can convery from hexadecimal to binary and back and forth.

After all, they're just reflections of the numbers that we represent using a specific set of symbols. In binary those symbols are more restrictive than in hexadecimal and therefore the symbolic representation is longer

Applications

CSS Colors

Funnily enough, if you've used a "HEX" value in HTML and CSS, you may already be loosely familiar with a similar scenario to what we walked through with the hexadecimal section.

For example, take the color #F33BC6 (a pinkish color). This color is a combination of 3 two-column hexadecimal numbers back-to-back. These numbers are:

F3, 3B, C6

They reflect the amount of red, green, and blue (respective) in said color. Because these numbers are two-column hexadecimal numbers, the highest a number can be to reflect one of these colors is 255 (which is FF in hexadecimal).

If you're unfamiliar with how red, green and blue can combine to make the colors we're familiar with (such as yellow, orange, purple, and much more), it might be worth taking a look at some of the color theory behind it. You can find resources on the topic on Wikipedia and elsewhere

These numbers, in decimal, are as follows:

HEX	Decimal
F3	243
3B	59
C6	196

And construct the amount of Red, Green, and Blue used to construct that color

Represents	HEX	Decimal
Red	F3	243
Green	3B	59
Blue	C6	196

Even without seeing a visual representation, you can tell that this color likely has a purple hue - since it has a high percentage of red and blue.

Text Encoding

While hexadecimal has much more immediately noticable application with colors, we started this post off with a question: "How does your computer know what letters to display on screen from only binary?"

The answer to that question is quite complex, but let's answer it in a very simple manor (despite missing a lot of puzzle pieces in a very "draw the owl" kind of way).

Let's take a real way that computers used to (and still ocationally do) represent letters internally: ASCII. ASCII is an older standard for representing letters as different numbers inside your computer. Take the following (simplified) chart:

When the user types "This", what the computer interprets (using ASCII) is 84, 104, 105, and 115 for T, h, i, and s respectively.

You might be wondering "Why is there a bunch of missing numbers"?

I've removed them to keep the examples simple, but many of them are for symbols (EG: #, /, and more) and some of them are for internal key commands that were used for terminal computing long ago that your computer now does without you noticing

It's also worth mentioning that ASCII, while there are more characters than what's presented here, was eventually replaced in various applications by Unicode and other text encoding formats as it lacks various functionality we expect of our machines today, such as emoji and non-latin symbols (like Kanji).

While I've used the above chart to reflect A as 65, it'd be more accurate to say that you computer interprets the symbol as 1000001 internally. This is again due to the fact that your computer must interpret every number and letter as binary.

Conclusion

While this has been only a high-level overview of how your computer interprets these non-decimal numbers (and some of their applications), it can provide some basic insights to what your computer is doing every time you type in a keystroke or see a color on screen. Under the hood everything is binary, and now you understand the introduction to how to convert binary to numbers you and I may understand better: to decimal!