}

When computers confuse margins: kilobytes and kibytes

2000/09/01 Mujika, Alfontso - Elhuyar Fundazioa Iturria: Elhuyar aldizkaria

Computing, and information technology in general, has had to create new concepts and terms to designate new hardware and software on computers and, in recent times, new services over the Internet. In the whirlwind of running, sometimes the new terms created have not been sufficiently precise. It's a quick job, a lazy job. Below we will see some of the issues that have arisen by some of these terms and their solution.

Computer Science Training

On one occasion there was a computer expert, who was still a new-born computer science (Informatics was born yesterday night, although many young people believe it is an old thing), who had to write large amounts of bytes, and who came up with an ingenious idea: electronic information is written in binary code, with only two digits: zeros and one. But the rest of the non-computer - and the computer also after turning off the computer - we use a 10-digit decimal metric system. But let's go to the thread: that skillful computer realized that 2 10 = 1024, that is, that it was approximately 1000, and so he said: "In the International System of Units they use the prefix "kg" to express 1000 (1 kilogram = 1000 grams). Therefore, I could also use the kilo prefix to express 1024, as from 1000 to 1024 there is a small difference. I make a small sin and give it right, slowly:

2 10 bytes = 1024 bytes = 1 kilobyte".

And so, he moved the idea to many other computer scientists. All right! is said and starts using the kilobyte. But the physicists, who knew it, put the black end: "Those computer gossip is merely encouragement! We have spent many years using the kilo prefix to express 1000 and now those mucous have not started to use to express 1024. Not to think!"

Grow and raise weeds

Computer scientists, as you have imagined, ignored the physical resins and continued to use the kilobyte to express 1024 bytes. At first, in the first 10 or 20 years, everything went well, since everyone who spoke of kilobytes knew how to talk about 1024 bytes. But at the beginning of the world of computers and computers that small and closed spread at full speed and computer vendors also had to communicate with physicists, engineers and street people. And physicists, engineers and normal street people know that the kilometer has 1000 meters and the kilogram 1000 grams.

Information accumulation systems have been developed very quickly and megabytes and gigabytes – and, lately, terabytes – are very common units. But information accumulation devices were not built according to the binary system and, for practical purposes, decimal arithmetic is more appropriate than binary arithmetic. Consequently, today, even if it seems a lie, we do not know how many bytes the megabyte has. For example, speaking of computer memory, most manufacturers use the megabyte drive to represent 2 20 = 1 048 576 bytes, while most manufacturers of information storage devices (disks) use the megabyte to represent 1 000 000 bytes. To confuse more margins, some local network designers have used the “megabit per second” unit to represent 1 048 576 bit/s, but all telecom engineers use the “megabit per second” unit to represent 1 000 000 bit/s. Not only that, since it is two few definitions of megabyte, a third has appeared that has 1 024 000 bytes and that is the one used to format the 3 1/2 inch (90 mm) floppy disk, 1.44 mb floppy that we all use.

Happy ending of the story

If you want a scramble, or a bytepil, it is not a joke and has become a real obstacle to the normalization and standardization of systems. Therefore, the resins physicists, the computer scientist, the anarino and the engineers, united, have wanted to eliminate the confusion and have put remedy. In December 1998, the International Electrotechnical Committee (IEC in French and Spanish, IEC in English, the highest international body responsible for the standardization of electrotechnology worldwide), with the collaboration of the Institute of Electrical and Electronic Engineers and the International Committee on Weights and Measures (SI International Unit), has the name of the multiple decisions of reverse processing for its use in information transmission.

These new prefixes for multiples of two are not part of the International System Unit SI, but are based on prefixes that are used in the SI system to represent the powers of 10. As can be seen in the box, the name of the new prefix SI derives from the name of the corresponding prefix: The first two letters of the SI prefix have been taken and the pair of letters 'two' has been added, which have been taken from 'binary'. Also, the symbol of each new prefix comes from the symbol of the corresponding SI prefix and the letter i has been added, taking back the word 'binar' (to match the symbols of the other prefixes of the multiples of Birs, the symbol 'Ki' has been approved to represent 2 10, capital letters, instead of the corresponding 'ki' symbol, based on the SI system).

Some illustrative examples of new savings and SI prefixes are:

  • 1 Kibit = 210 bit = 1024
  • 1 bit kilobit = 103 bit = 1000 bit
  • 1 Mebibyte 1 MiB = 220 B = 1 048 576 B
  • 1 megabyte 1MB = 106 B = 1 000 000 000 000 B
  • 1 Gigibyte 1 GiB = 230 B = 1 073 741 824 B
  • 1 gigabyte 1GB = 103 B = 1 000 000 000 000 000 B

Decisions are relatively new, as they were published last year, and will take a few years until they are fully used anywhere and at any time, as with all new units. It can be said, therefore, that the chandro that computer scientists have made to physicists a few years ago is recovering. But the polemics between computer scientists and physicists do not end there. For example, for physicists the symbol of the candle unit is B (in most cases its submultiple is used: dB, decibel), while for computer users the letter B is the symbol of the byte unit. There they have another way of confrontation that still has no solution.

Gai honi buruzko eduki gehiago

Elhuyarrek garatutako teknologia