WEBVTT - TechStuff Tidbits: Bits and Bytes 0:00:04.400 --> 0:00:07.800 Welcome to tex Stuff, a production from I Heart Radio. 0:00:12.119 --> 0:00:14.880 Hey there, and welcome to tech Stuff. I'm your host, 0:00:15.040 --> 0:00:18.000 Johan Strickland. I'm an executive producer with I Heart Radio. 0:00:18.040 --> 0:00:20.640 And how the tech are you? It's time for a 0:00:20.760 --> 0:00:23.400 tech Stuff tidbits. And I thought it would be a 0:00:23.440 --> 0:00:26.240 good idea to do an episode to talk about bits 0:00:26.320 --> 0:00:28.960 and bytes, largely because I think a lot of folks 0:00:29.000 --> 0:00:32.560 can confuse the two, include myself in that, and to 0:00:32.680 --> 0:00:37.640 be fair, it is confusing, like even when you're in 0:00:37.680 --> 0:00:41.600 the computer science field, this can get confusing. It's totally understandable. 0:00:41.840 --> 0:00:44.960 So we talk about data transfer rates in terms like 0:00:45.120 --> 0:00:49.080 megabits or gigabits per second, but we talk about data 0:00:49.159 --> 0:00:53.680 storage in terms of gigabytes or terabytes. And then we 0:00:53.760 --> 0:00:58.320 talk about you know, memory in terms of megabytes or gigabytes, 0:00:58.560 --> 0:01:01.360 and we mean different negabytes and gigabytes that we do 0:01:01.480 --> 0:01:04.240 with data storage. So it's not hard to get this 0:01:04.280 --> 0:01:08.520 stuff mixed up. But let's start with the bit. It's 0:01:08.560 --> 0:01:12.160 the easiest one to understand. So in computer terms, a 0:01:12.240 --> 0:01:19.040 bit is the smallest unit of information. So in computer data, 0:01:19.280 --> 0:01:21.880 you know analogies, you would call a bit the same 0:01:21.920 --> 0:01:25.399 thing as an atom. As like the building block for 0:01:25.520 --> 0:01:31.040 computer information, and a bit is a binary digit. It's 0:01:31.080 --> 0:01:36.280 how we, you know, can talk about individual pieces of 0:01:36.400 --> 0:01:39.440 data and we represent it as being either a zero 0:01:39.560 --> 0:01:42.440 or a one. It's base two. It's a zero or 0:01:42.480 --> 0:01:44.440 a one. And I often say you can think of 0:01:44.480 --> 0:01:47.320 a bit kind of like an on or off switch, 0:01:47.560 --> 0:01:49.920 and you can say like, well, one means the switches 0:01:50.000 --> 0:01:54.480 on and zero means the switches off. John Wilder, two 0:01:54.560 --> 0:02:00.200 key mathematician, suggested the term bit back in n So 0:02:00.240 --> 0:02:04.080 we're talking about the very early days of modern computer science. 0:02:04.080 --> 0:02:06.720 I mean a lot of groundwork had been laid by 0:02:06.800 --> 0:02:10.960 people like you know, Charles Babbage and Ada Lovelace. But 0:02:11.600 --> 0:02:15.440 the late forties is really where we start getting the 0:02:15.639 --> 0:02:20.400 foundation of modern computer science, and computers had been around 0:02:20.440 --> 0:02:24.079 a little bit, byn not by a lot. We were 0:02:24.120 --> 0:02:28.040 pretty young, um, and a lot of the computer technology 0:02:28.080 --> 0:02:30.960 actually evolved from earlier machines that were meant to do 0:02:31.040 --> 0:02:36.280 everything from count ballots to guide a loom when weaving 0:02:36.480 --> 0:02:40.120 a specific pattern. The evolution of the bit involves stuff 0:02:40.160 --> 0:02:41.960 like punch cards. But to get into all of that 0:02:42.000 --> 0:02:45.000 would be a little bit too much for a Tidbits episode. 0:02:45.520 --> 0:02:48.680 So two key coined this term, but it was Claude 0:02:48.680 --> 0:02:52.160 Shannon who really popularized it in his work titled A 0:02:52.200 --> 0:02:56.919 Mathematical Theory of Communication. He credited credited two Key in 0:02:57.000 --> 0:02:59.920 that work, so Shannon didn't try and pass this all 0:03:00.000 --> 0:03:02.720 phys his own idea. I think that's awesome because I 0:03:02.800 --> 0:03:04.680 don't see that a lot, you know, I see people 0:03:05.200 --> 0:03:08.120 using terms and not indicating that, hey, someone else actually 0:03:08.120 --> 0:03:11.400 thought this up. That wasn't Claude Shannon's style. Shannon was 0:03:11.480 --> 0:03:13.959 quick to credit to Key with coming up with the 0:03:14.560 --> 0:03:19.680 idea anyway. Shannon laid out the that a device capable 0:03:19.840 --> 0:03:24.560 of two stable positions or states such as off and on. 0:03:24.760 --> 0:03:28.400 There's the off state and the on state. Well, something 0:03:28.440 --> 0:03:32.640 like that can store one bit of information, and that 0:03:32.720 --> 0:03:36.920 this meant for in number of such devices you can 0:03:37.000 --> 0:03:40.840 store in bits of information. So, in other words, if 0:03:40.880 --> 0:03:44.720 you have twenty switches right, and each of the switches 0:03:44.760 --> 0:03:47.600 has an on or off position, you can store up 0:03:47.600 --> 0:03:51.800 to twenty bits with that system. From there, Shannon dives 0:03:51.840 --> 0:03:54.120 into how this approach can be used to communicate on 0:03:54.160 --> 0:03:58.680 a computational basis. The paper itself is free to read online. 0:03:58.760 --> 0:04:01.160 You can find it again. You just just search for 0:04:01.200 --> 0:04:04.960 the title, which was a mathematical theory of communication. It'll 0:04:04.960 --> 0:04:07.920 pop right up and you can read it. Uh, it's 0:04:08.040 --> 0:04:11.160 a technical document, and honestly, it's the kind of paper 0:04:11.160 --> 0:04:13.280 where I need to have a separate tab open so 0:04:13.360 --> 0:04:16.960 I can look up terms and meanings just to try 0:04:17.000 --> 0:04:20.400 to keep up. And even that is being generous. It 0:04:20.400 --> 0:04:23.760 it is. Uh, someone in computer science totally makes sense 0:04:23.800 --> 0:04:27.120 to them, like, no no brainer. For someone like me 0:04:27.200 --> 0:04:30.200 with my background in English literature, it requires a bit 0:04:30.200 --> 0:04:32.919 more homework on my part. But it is a truly 0:04:32.960 --> 0:04:36.680 fascinating and foundational piece of work in the computer science discipline. 0:04:37.960 --> 0:04:40.520 But then, what can you represent if you have just 0:04:40.880 --> 0:04:45.440 a single bit with a switch that's off or on. Well, 0:04:45.839 --> 0:04:49.200 with just two states, you can't really represent anything terribly 0:04:49.360 --> 0:04:54.240 useful for communication. Uh, you could do yes, no, but 0:04:54.320 --> 0:04:57.320 that's it. Like you couldn't form a question. You could 0:04:57.360 --> 0:05:03.520 just maybe given answer that's very very very simple, But 0:05:03.600 --> 0:05:06.120 you couldn't really process information with a single bit, like 0:05:06.360 --> 0:05:09.960 a processors that can only handle one bit would be useless. 0:05:10.000 --> 0:05:12.480 So let's look at what happens when we have more 0:05:12.600 --> 0:05:16.080 bits in our disposal. Well, each bit again has two 0:05:16.160 --> 0:05:19.560 potential states off on zero, one. But if you have 0:05:19.600 --> 0:05:24.359 two bits together, well, then you can get a shaven haircut. Sorry, 0:05:24.400 --> 0:05:27.640 that's a very ill dated dad joke. If you happen 0:05:27.760 --> 0:05:31.640 to know the whole shaven a haircut two bits, good 0:05:31.640 --> 0:05:34.920 for you. You might have appreciated that very bad joke 0:05:35.000 --> 0:05:37.240 I made. No. No. If you have two bits, you 0:05:37.320 --> 0:05:42.120 technically can represent four states with those two bits, right, 0:05:42.400 --> 0:05:45.359 you can have zero zero, that's the first one. You 0:05:45.360 --> 0:05:48.680 can have zero one, you could have one zero, or 0:05:48.680 --> 0:05:50.600 you could have one one. So with two bits you 0:05:50.640 --> 0:05:56.160 can represent four things. Well, what if you had four bits, Well, 0:05:56.240 --> 0:05:59.720 that means you could represent sixteen different outcomes and they 0:05:59.720 --> 0:06:03.840 would range from zero zero, zero zero to one one 0:06:03.839 --> 0:06:06.920 one one, and there will be sixteen different ones. If 0:06:06.920 --> 0:06:09.480 you had eight bits, you could represent up to two 0:06:09.640 --> 0:06:14.680 hundred fifty six versions or outcomes. So the easy way 0:06:14.720 --> 0:06:17.000 to represent this is to take the number two. That 0:06:17.080 --> 0:06:19.600 represents the number of states that each bit can have. 0:06:20.200 --> 0:06:22.360 You know, a zero or a one, that's two states. 0:06:22.600 --> 0:06:24.840 So you take the number two and you raise that 0:06:24.880 --> 0:06:28.000 too to a power equivalent to the number of bits 0:06:28.080 --> 0:06:31.720 you're talking about. So eight bits is the same as 0:06:31.720 --> 0:06:35.080 saying two to the eighth power or two hundred fifty 0:06:35.200 --> 0:06:40.040 six potential values. So this means that as you double bits, 0:06:40.120 --> 0:06:41.880 you are you know, or as you increase bits, you 0:06:41.920 --> 0:06:46.320 are logarithmically increasing the number of potential states. So if 0:06:46.360 --> 0:06:50.360 we double eight bits to get sixteen bits, that doesn't 0:06:50.400 --> 0:06:55.120 mean that we double two hundred fifty six to twelve. No, no, no, no, 0:06:55.520 --> 0:06:59.120 it is two to the power of sixteen. That that 0:06:59.279 --> 0:07:02.800 is sixty five thousand ft six. So you see that 0:07:02.920 --> 0:07:07.240 as you add bits to a system, you dramatically increase 0:07:07.440 --> 0:07:09.600 the number of states. In fact, every time you add 0:07:09.600 --> 0:07:14.760 a bit two systems capability of handling bits, you double 0:07:15.480 --> 0:07:20.280 the number of of potential values you can represent. So 0:07:20.720 --> 0:07:24.080 this is what I meant by a logarithmic increase. All right, 0:07:24.200 --> 0:07:27.360 So once you go with binary digits, you start to 0:07:27.400 --> 0:07:29.440 look at how many bits you need to do whatever 0:07:29.480 --> 0:07:31.920 it is you need to do. So let's say that 0:07:32.000 --> 0:07:35.240 you want to start off just by representing the Latin alphabet, right, 0:07:35.680 --> 0:07:38.880 You want to be able to use bits to designate 0:07:39.080 --> 0:07:42.160 letters of the alphabet and say this combination of bits 0:07:42.240 --> 0:07:45.080 means A, this one means B. Well, if you're just 0:07:45.120 --> 0:07:47.360 looking at the number of letters in the Latin alphabet, 0:07:47.680 --> 0:07:51.720 we would need at least twenty six values. Right, we 0:07:51.720 --> 0:07:56.000 would need twenty six different combinations in order to represent 0:07:56.680 --> 0:08:00.880 the just the basic alphabet. Full were bits would let 0:08:00.960 --> 0:08:05.200 us sixteen values, so that's not enough, But five bits 0:08:05.240 --> 0:08:07.760 would give us thirty two as because two to the 0:08:07.800 --> 0:08:12.360 fifth power is thirty two. So with five bits we 0:08:12.400 --> 0:08:15.120 could represent all the letters of the alphabet, and we'd 0:08:15.160 --> 0:08:17.840 have a couple of values left over where we could 0:08:17.840 --> 0:08:22.800 represent simple punctuation. However, we wouldn't be able to have 0:08:23.080 --> 0:08:25.760 upper and lower case letters. All letters would have to 0:08:25.760 --> 0:08:28.640 be the same case because each case would be a 0:08:28.760 --> 0:08:32.640 state or a value of its own. So capital J 0:08:33.000 --> 0:08:36.280 and a lower case J would each require their own designations, 0:08:36.760 --> 0:08:38.560 and we don't have enough bits to do that. If 0:08:38.559 --> 0:08:40.560 we have thirty two, we we would have have to 0:08:40.600 --> 0:08:42.719 have at least fifty two in order to do that, 0:08:43.200 --> 0:08:46.040 and we don't have that. We've got thirty two plus. 0:08:46.080 --> 0:08:48.679 We wouldn't be able to represent any numerals, or at 0:08:48.720 --> 0:08:50.800 least not all of them. Would just thirty two bits, 0:08:51.120 --> 0:08:53.679 unless we were to sacrifice some letters of the alphabet, 0:08:54.000 --> 0:08:56.440 because otherwise the alphabet takes up too many of the 0:08:56.480 --> 0:09:01.160 states or values. Now, the term by began to pop 0:09:01.280 --> 0:09:04.439 up a little bit around this time to describe the 0:09:04.520 --> 0:09:09.200 number of bits engineers were using to represent a character set. So, 0:09:09.280 --> 0:09:12.840 for example, if we used five bits to represent all 0:09:12.920 --> 0:09:15.520 the characters in our set, which I mean again, we 0:09:15.520 --> 0:09:18.560 would be limited to thirty two characters, then we would 0:09:18.600 --> 0:09:23.959 naturally refer to five bits as a byte in our system. 0:09:24.000 --> 0:09:26.760 And you've probably heard that eight bits are a byte. 0:09:27.080 --> 0:09:30.440 Well they are now, but in the early days, what 0:09:30.559 --> 0:09:33.040 you're referred to as a byte depended upon the system 0:09:33.160 --> 0:09:36.600 architecture you were working with at the time, so the 0:09:36.679 --> 0:09:41.280 bite was not always in forever on men. Uh, you know, 0:09:41.400 --> 0:09:44.480 eight bits, that's not the way that worked. There were 0:09:44.520 --> 0:09:48.560 five bit bytes, six bit bytes, seven bit bytes. Uh. 0:09:48.600 --> 0:09:53.160 These were all kind of hashing out over time as 0:09:53.240 --> 0:09:57.160 various companies were building out computer systems. Also in the 0:09:57.160 --> 0:09:59.880 early days of computing, designers created machines that had to 0:10:00.040 --> 0:10:03.760 for an instruction set. Architectures, and some systems used a 0:10:03.840 --> 0:10:08.000 five bit sized word, which is a collection of bits 0:10:08.200 --> 0:10:12.240 that becomes the native unit of storage. By storage, we're 0:10:12.280 --> 0:10:14.880 not just talking about storing data like in a hard drive. 0:10:15.400 --> 0:10:18.440 We're really talking about storage in the sense of a 0:10:18.440 --> 0:10:20.880 computer has to be able to hold onto a certain 0:10:20.880 --> 0:10:25.400 amount of information in order to process the information, you know, 0:10:25.440 --> 0:10:28.160 to execute some sort of operation on the data. So 0:10:28.280 --> 0:10:32.120 you're you're essentially talking about was the processor's capacity, how 0:10:32.200 --> 0:10:36.160 much information can it hold in order to do operations 0:10:36.200 --> 0:10:39.920 on that data. So smaller word size means you are 0:10:39.960 --> 0:10:44.680 working with smaller amounts of information that the processor can handle, 0:10:44.760 --> 0:10:47.040 and it limits what your processor can do and thus 0:10:47.080 --> 0:10:50.160 limits what your computer can do. So when we talk words, 0:10:50.400 --> 0:10:54.880 we're talking about stuff like CPU registers, which temporarily store 0:10:54.920 --> 0:10:58.240 small pieces of information while the CPU executes some sort 0:10:58.240 --> 0:11:01.600 of process on that data. Some early systems used five 0:11:01.679 --> 0:11:05.440 bit words, some used six bit words. Uh, and then 0:11:05.480 --> 0:11:08.480 they grew very quickly from there because that was so 0:11:08.600 --> 0:11:12.000 limited that you couldn't do much with them. All Right, 0:11:13.120 --> 0:11:15.600 we were just getting started. We're gonna take a quick break. 0:11:15.640 --> 0:11:17.960 When we come back, we will continue to talk about 0:11:18.000 --> 0:11:28.559 bits and bites. Okay, So in the early days of computing, 0:11:28.760 --> 0:11:35.000 the technological limitations would determine word length. So let's let's 0:11:35.040 --> 0:11:38.360 review really quickly. A bit is a single unit of information. 0:11:38.400 --> 0:11:41.760 It's either a zero or a one. A bite is 0:11:41.800 --> 0:11:45.319 a consecutive group of bits, which we used to represent characters, 0:11:45.880 --> 0:11:49.120 and a word refers to a consecutive number of bits 0:11:49.240 --> 0:11:53.600 or bites, used primarily for CPU registers, and you can 0:11:53.640 --> 0:11:55.800 think of word size as being an indication of how 0:11:55.880 --> 0:11:59.640 much information on computers CPU can handle for individual operations, 0:11:59.760 --> 0:12:04.599 large word size means the computer can handle bigger operations effectively. 0:12:05.760 --> 0:12:08.679 By the nineteen fifties, various companies began using a character 0:12:08.720 --> 0:12:11.040 set called b c D, which had you know at 0:12:11.120 --> 0:12:13.720 least forty eight characters in it, and to encode for 0:12:13.760 --> 0:12:15.920 the eight characters you would need to have at least 0:12:16.080 --> 0:12:19.440 six bits, So the six bit bite became kind of 0:12:19.480 --> 0:12:22.679 a standard for a while. By the time IBM was 0:12:22.720 --> 0:12:25.760 ready to introduce the System three sixty, the company had 0:12:25.800 --> 0:12:30.160 gravitated towards an eight bit size for bites. Now, technically 0:12:30.600 --> 0:12:33.640 the three sixty could get by with just seven bits 0:12:33.679 --> 0:12:36.240 to represent all the characters that it was going to use, 0:12:36.960 --> 0:12:41.200 but programming is way easier if you're dealing with bites 0:12:41.400 --> 0:12:45.600 and words that are based on powers of two. Seven 0:12:45.679 --> 0:12:48.839 is not a power of two, but eight is, so 0:12:49.160 --> 0:12:52.120 bumping up the bite size from seven bits to eight 0:12:52.120 --> 0:12:54.839 bits made more practical sense, and it also meant you 0:12:54.920 --> 0:12:57.920 had two hundred fifty six values to play with instead 0:12:57.920 --> 0:12:59.800 of a hundred twenty eight, which is what you would 0:12:59.800 --> 0:13:02.199 get if you were using seven bits to a byte. 0:13:02.640 --> 0:13:06.400 IBMS move would end up creating the foundation for bites 0:13:06.440 --> 0:13:09.640 moving forward, though it didn't catch on immediately. So by 0:13:09.679 --> 0:13:12.720 the time I was a kid learning about personal computers 0:13:12.720 --> 0:13:16.040 in the late seventies and early eighties, a byte was 0:13:16.200 --> 0:13:19.360 pretty firmly established as being eight bits, and in fact, 0:13:19.440 --> 0:13:22.320 I don't remember ever seeing anything that suggested that had 0:13:22.360 --> 0:13:24.600 not always been the case. So I walked away with 0:13:24.600 --> 0:13:27.240 the impression that, you know, a bit was always a 0:13:27.360 --> 0:13:29.880 zero or a one, and a byte had always been 0:13:29.960 --> 0:13:33.520 eight bits, but the eight bit byte really wasn't standardized 0:13:33.559 --> 0:13:37.360 until say, the early nineteen seventies. So you've got your bit, 0:13:37.920 --> 0:13:40.959 you've got your bite, which is eight bits and super quick. 0:13:41.120 --> 0:13:44.880 We should reference what prefixes like kilo, mega, giga, terra 0:13:45.240 --> 0:13:47.720 or killa. This is that's the way most people say it, 0:13:47.760 --> 0:13:50.440 and when they're talking about bytes and bits what those mean, 0:13:50.920 --> 0:13:54.360 And they come from metric prefixes, but bits and bytes 0:13:54.640 --> 0:13:57.440 are not metric units, and when it comes to bites 0:13:57.480 --> 0:14:01.880 in particularly gets real confusing. So kilo or killa means 0:14:02.040 --> 0:14:07.400 one thousand, Mega means million, Giga means billion, Tera means trillion, 0:14:07.520 --> 0:14:10.559 and you can go further. I mean, petta would be quadrillion, 0:14:11.040 --> 0:14:13.400 exa would be quintillion. So if you hear something like 0:14:13.440 --> 0:14:17.120 an exabyte, they're talking about the quintillion bytes and then 0:14:17.160 --> 0:14:19.440 it keeps on going up from there. But for the 0:14:19.480 --> 0:14:22.600 average person, terra is kind of where we max out 0:14:22.640 --> 0:14:25.600 when we're talking about modern personal computers, like a terabyte 0:14:25.640 --> 0:14:28.200 hard drive, that kind of thing. Now, let's talk about 0:14:28.840 --> 0:14:32.840 data transfer speeds and computer storage and computer memory and 0:14:32.880 --> 0:14:35.280 why we use terms like gigabit or gigabyte and what 0:14:35.320 --> 0:14:38.920 does actually mean. So when we're talking about transmission speeds, 0:14:39.280 --> 0:14:43.920 we are framing the discussion in terms of bits per second. Uh. 0:14:43.960 --> 0:14:49.640 This also can kind of get a little confused with bandwidth. UH, 0:14:49.680 --> 0:14:53.600 that's pretty easy to confuse with transmission speed. Bandwidth describes 0:14:53.640 --> 0:14:56.440 the capacity of a network. UH. In other words, the 0:14:56.440 --> 0:14:59.560 amount of data that network can handle or transmit at 0:14:59.600 --> 0:15:05.440 any give and time. And networks are not infinitely capable 0:15:05.520 --> 0:15:08.360 of handling data, they do have a cap This is 0:15:08.400 --> 0:15:09.880 one of those things that I s p s like 0:15:10.000 --> 0:15:13.320 to reference when they're talking about having to charge people 0:15:13.320 --> 0:15:16.480 in arm in a leg to access the Internet because 0:15:16.960 --> 0:15:21.040 there are capacity limits that a network will hit. Uh, 0:15:21.160 --> 0:15:22.960 most of the time we don't get to the full 0:15:23.000 --> 0:15:26.800 capacity limit. But that's still the story we get told whenever, 0:15:27.040 --> 0:15:30.120 whenever it's time to pay the bill. So um, there's that. 0:15:30.200 --> 0:15:33.120 But transmission speed is literally the rate at which data 0:15:33.200 --> 0:15:36.000 crosses from one point in a network to another, and 0:15:36.040 --> 0:15:38.480 that depends on a whole bunch of stuff, Like it 0:15:38.520 --> 0:15:41.600 can depend upon the distance between the two points, right, 0:15:41.640 --> 0:15:45.360 Like if you are transferring data from a computer that's 0:15:45.360 --> 0:15:47.760 on the opposite side of the world where from where 0:15:47.800 --> 0:15:51.560 you are, that's going to affect the transmission speed as 0:15:51.560 --> 0:15:54.760 opposed to like a computer that's you know, a mile away. 0:15:55.600 --> 0:15:57.920 Then like the kinds of connections, like what kinds of 0:15:57.960 --> 0:16:00.560 wires are connecting you as a copper is it fiber? 0:16:01.280 --> 0:16:05.720 You know, that sort of stuff also determines the transmission speed. 0:16:06.240 --> 0:16:08.840 There are lots of other elements that do too, but 0:16:09.320 --> 0:16:12.520 ultimately you figure out, you know, what your transmission speed is. 0:16:12.760 --> 0:16:16.720 Here in the United States, the Federal Communications Commission currently 0:16:16.760 --> 0:16:20.480 defines broadband internet speed as twenty five megabits per second down. 0:16:21.040 --> 0:16:24.600 That means the that's the data transfer rate that applies 0:16:24.600 --> 0:16:27.720 to information that's coming from the Internet to your computer. 0:16:28.440 --> 0:16:31.640 That that if it's twenty five megabits per second or faster, 0:16:32.200 --> 0:16:35.040 then that means you have broadband access at least on 0:16:35.080 --> 0:16:39.080 the down download side, and three megabits per second up. 0:16:39.480 --> 0:16:41.520 So this would be the speed at which your computer 0:16:41.920 --> 0:16:44.680 sends data back up to the Internet. So if you 0:16:44.720 --> 0:16:47.720 have twenty five megabits down and three megabits up, that 0:16:47.800 --> 0:16:51.200 counts as broadband in the United States. Uh. And since 0:16:51.280 --> 0:16:57.080 mega means millions, that means million bits per second downloading, 0:16:57.440 --> 0:17:01.080 three million bits per second uploading. By the way, there 0:17:01.120 --> 0:17:03.720 are a lot of folks out there, including myself, who 0:17:03.760 --> 0:17:06.840 say this definition is way too low and we should 0:17:06.880 --> 0:17:10.600 have a higher standard to qualify for broadband designation. And 0:17:10.640 --> 0:17:14.000 that is important. It's not just semantics. It's important because 0:17:14.040 --> 0:17:18.440 there are various government initiatives that are dedicated to extending 0:17:18.520 --> 0:17:23.280 broadband service too underserved regions and populations in the United States, 0:17:23.840 --> 0:17:26.800 because people have recognized access to the Internet is one 0:17:26.840 --> 0:17:30.639 of the most important elements to participating in modern society, 0:17:30.960 --> 0:17:35.720 particularly during a pandemic. And so if you define broadband 0:17:35.800 --> 0:17:39.000 as a very low standard, you're you're not really helping 0:17:39.000 --> 0:17:41.919 out people with these programs to get access to that 0:17:42.119 --> 0:17:44.480 very low standard, like companies are going to do the 0:17:44.520 --> 0:17:47.080 bare minimum they need to do in order to get 0:17:47.119 --> 0:17:51.560 that access to those people. So you could argue that 0:17:51.600 --> 0:17:56.560 this definition would keep people at a technological disadvantage and 0:17:56.600 --> 0:18:00.280 that we really should uh change the definition and to 0:18:00.440 --> 0:18:05.800 be more reflective of what broadband really is. Anyway, transfer 0:18:05.840 --> 0:18:10.640 speeds are all in bits per second, and the prefixes kilo, mega, giga, 0:18:10.720 --> 0:18:13.320 and so on are very straightforward. Kill a bit is 0:18:13.359 --> 0:18:15.600 a thousand bits, so I kill a bit per second. 0:18:16.960 --> 0:18:19.080 Transfer speed would be terrible, but it would be a 0:18:19.160 --> 0:18:21.560 thousand bits per second, and a mega bit is a 0:18:21.560 --> 0:18:25.399 million bits, so very easy to follow. But it is 0:18:25.400 --> 0:18:27.679 a very different story when we talk about bites, and 0:18:27.720 --> 0:18:30.359 it's confusing, is all heck to a lot of folks, 0:18:30.359 --> 0:18:33.840 including folks in computer science. All right, So in the 0:18:33.840 --> 0:18:36.919 early days, there was this real need to stick to 0:18:37.119 --> 0:18:41.440 powers of two when you were talking about bites. Again, 0:18:42.440 --> 0:18:44.199 this dates all the way back to the introduction of 0:18:44.240 --> 0:18:48.520 the IBM system three sixty where executed IBM, we're saying no, no, no, 0:18:48.640 --> 0:18:52.240 let's let's just deal with powers of two. It simplifies things. 0:18:52.720 --> 0:18:57.720 Otherwise stuff breaks down. So rather than describe one thousand bites, 0:18:57.800 --> 0:19:00.240 you know that a killer bite is a thousand bights. 0:19:01.240 --> 0:19:04.320 It was more elegant within the computer science model to 0:19:04.400 --> 0:19:08.720 describe a kilo bite as one thousand twenty four bites. 0:19:09.240 --> 0:19:11.520 A thousand twenty four is the same thing as two 0:19:11.560 --> 0:19:15.800 to the tenth power, but two to the tenth power 0:19:15.960 --> 0:19:19.159 or one thousand twenty four. There there's no easy naming 0:19:19.200 --> 0:19:23.720 convention to use for to describe one thousand twenty four bites, 0:19:24.200 --> 0:19:27.080 and the computer science world kind of appropriated kill a 0:19:27.160 --> 0:19:30.320 bite to describe it, because a thousand twenty four is 0:19:30.480 --> 0:19:33.160 kind of like a thousand if you squint your eyes 0:19:33.200 --> 0:19:37.320 a little. From a computational standpoint, sticking with powers of 0:19:37.359 --> 0:19:42.160 two made things easier. From a semantic standpoint, it done 0:19:42.200 --> 0:19:44.480 mess things up because I kill a bit is a 0:19:44.520 --> 0:19:47.240 thousand bits, But to kill a bite at least originally 0:19:47.600 --> 0:19:52.719 was a thousand twenty four bites bud weight. It'll get worse, 0:19:52.960 --> 0:20:04.360 I'll explain after we come back from this quick break. Okay, 0:20:04.840 --> 0:20:08.159 a kilobyte is a thousand twenty four bytes because we 0:20:08.200 --> 0:20:10.320 wanted to stick to that power of two things. Well, 0:20:10.359 --> 0:20:14.200 then we get to megabyte. Well, mega means million, so 0:20:14.520 --> 0:20:18.440 megabyte should mean one million bytes. But in those same 0:20:18.480 --> 0:20:21.439 little areas of computer science, particularly those dealing with like 0:20:22.080 --> 0:20:25.920 computer memory, that kind of stuff, a megabyte was really 0:20:25.960 --> 0:20:29.680 seen as actually being one million, forty eight thousand, five 0:20:29.760 --> 0:20:34.480 hundred seventy six bytes. And you might say, what, what why? 0:20:34.560 --> 0:20:37.720 Well again, it's those powers of two, So a kilobyte 0:20:37.800 --> 0:20:39.920 was two to the tenth power. A megabyte was two 0:20:40.000 --> 0:20:43.920 to the twenty power or one thousand, twenty four squared. 0:20:44.760 --> 0:20:47.399 But hey, one million, forty eight thousand, five d seventy 0:20:47.440 --> 0:20:49.159 six bytes is kind of hard to say, right, So 0:20:49.240 --> 0:20:52.880 let's just call it a megabyte, right, It's called a megabyte. 0:20:53.240 --> 0:20:57.840 Who's gonna care. By the late nineties, the International Electrochemical 0:20:57.840 --> 0:21:01.119 Commission had had enough of this nonsense because it was 0:21:01.240 --> 0:21:04.679 causing tons of confusion. I mean, the computer science world 0:21:04.720 --> 0:21:07.639 was going all humpty dumpty on the rest of us. Uh, 0:21:08.000 --> 0:21:11.199 if you don't understand that reference, Okay, And Alice and 0:21:11.320 --> 0:21:14.119 through the looking glass there's this encounter she has with 0:21:14.200 --> 0:21:16.840 Humpty dumpty, you know, the egg that sat on the wall, 0:21:17.400 --> 0:21:20.880 and Humpty Dumpty says words mean whatever he wants them 0:21:20.880 --> 0:21:23.240 to mean. He says, like, you know, the only question 0:21:23.280 --> 0:21:25.679 is who who is to be the master the words 0:21:25.800 --> 0:21:28.639 or me? And I'm not letting the words push me around. 0:21:28.880 --> 0:21:30.960 So when I use words, they mean exactly what I 0:21:31.000 --> 0:21:33.159 want them to mean. That's kind of what the computer 0:21:33.200 --> 0:21:34.959 science world was doing to the rest of us, And 0:21:35.000 --> 0:21:37.879 the brave among us said, Yo, you can't do that. 0:21:37.960 --> 0:21:42.680 Words mean things. So anyway, the I e C made 0:21:42.680 --> 0:21:48.480 a recommendation that KILLO, mega, giga, etcetera. Would mean the 0:21:48.520 --> 0:21:51.399 same thing they mean in metric systems. So in other words, 0:21:51.560 --> 0:21:55.359 if you use the word megabyte, you meant one million bytes. 0:21:55.880 --> 0:21:58.119 And if you wanted to go to the power of 0:21:58.160 --> 0:22:00.600 two route like if you if you really wanted to 0:22:00.640 --> 0:22:06.280 call one million, forty eight thousand bytes something, the I 0:22:06.400 --> 0:22:09.240 e C said, well, don't use megabyte. That's confusing. Will 0:22:09.320 --> 0:22:12.600 create a new designation. Call it a membe bite, m 0:22:12.600 --> 0:22:16.760 E B I byte b y t E. So the 0:22:16.800 --> 0:22:18.920 one thousand, twenty four version that wouldn't be called a 0:22:18.960 --> 0:22:22.040 kilobyte anymore, that'd be called a kippi byte. And they 0:22:22.040 --> 0:22:24.639 also went ahead and said the one thousand to twenty 0:22:24.680 --> 0:22:26.920 four to the third power or two to the thirty 0:22:27.119 --> 0:22:29.960 power would no longer be a gigabyte. That would be 0:22:30.000 --> 0:22:33.080 a gimby byte, and one thousand twenty four to the 0:22:33.119 --> 0:22:36.560 fourth power or two to the power would be a 0:22:36.600 --> 0:22:40.040 tabby byte, not a terabyte, and so on, and so 0:22:40.320 --> 0:22:43.119 if you said gigabyte, you meant a billion bytes, and 0:22:43.280 --> 0:22:47.000 terabyte would be a trillion bytes. This was meant to 0:22:47.000 --> 0:22:51.040 clarify things so that everyone knew what people were talking 0:22:51.040 --> 0:22:55.119 about when they were using a specific designation, and that 0:22:55.160 --> 0:22:57.760 would clear everything up, except the computer science world at 0:22:57.840 --> 0:23:01.919 large kind of ignored the suggestion. So there remains this 0:23:02.080 --> 0:23:07.040 use of the terminology that, depending upon the context, will 0:23:07.080 --> 0:23:10.879 mean one number of bites or a different number of bites. 0:23:11.160 --> 0:23:14.560 Like if you're talking about RAM, for example, you're really 0:23:14.600 --> 0:23:16.920 referring to the power of two version the base to 0:23:17.119 --> 0:23:21.080 description in other words, like the one twenty four for 0:23:21.560 --> 0:23:24.160 kill a byte. But if you're talking about hard drives, 0:23:24.520 --> 0:23:27.280 while you're typically talking about the base ten version, because 0:23:27.320 --> 0:23:31.800 these days hard drives when they're marketed, are marketed toward that. 0:23:32.280 --> 0:23:36.240 So a five hundred gigabyte hard drive is supposed to 0:23:36.240 --> 0:23:40.240 be five hundred billion bytes, although it's usually slightly off 0:23:40.280 --> 0:23:42.520 from that, but it's supposed to be in that neighborhood, 0:23:42.520 --> 0:23:45.480 and it's not you know, the the power of two 0:23:45.680 --> 0:23:49.760 variation of of that. So yeah, it's all clear as mud. Right, 0:23:50.280 --> 0:23:52.879 So kill a bit is different from a kilobyte, and 0:23:52.960 --> 0:23:56.320 sometimes a kilobyte is different from a different kilobyte depending 0:23:56.359 --> 0:24:01.840