Talk:Kibibyte

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Computing  
WikiProject icon This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
 ???  This article has not yet received a rating on the project's quality scale.
 ???  This article has not yet received a rating on the project's importance scale.
 

Unnamed section[edit]

Discussion about centralization took place at Talk:Binary prefix.

Kilobyte means 1024 bytes. What were the IEC thinking making another term to mean 1024 bytes, they should have made a term to mean 1000 bytes and shamed drive manufacturers into using that. Then there wouldn't be any confusion and there would be two clearly defined words. But now kilobyte is still just as ambiguous and we have this other word that no-one really uses because it's redundant.115.188.135.11 (talk) 03:22, 13 December 2012 (UTC)

How do you pronounce "kibibyte"? Is it "Kee Bee Bite" or "Kib Bee Bite" or something else? Perhaps a pronounciation guide is advised.

No one actually pronounces it. If and when people start using this, a common pronunciation will develop.

154.20.205.22 05:04, 13 September 2005 (UTC)

KiBi stands for Kilo Binary, that may help. It's probably not really meant to pronounced in its abbreviated form. There will be a difference anyway depending on what's your native tongue. "Euro" is pronounced differently in different states of Europe as well. The same applies to Kilo, Mega, Giga etc. which are pronounced quite different by English and German speakers. --82.141.58.95 23:56, 30 November 2005 (UTC)
I'm pretty sure the standard says that you pronounce the first syllable how you would if it where kilo (&c), and the next syllable as be. Therefore, kibi would be /kiːbiː/ (or /kibi/ in the US). That's the way I'd naturally pronounce it anyway.
Where are you getting your IPA symbols? I'm from the US, and "kilo" is pronounced with a short "i". That makes it /'kɪ•bi:/. (Please don't forget to indicate stress.) Bostoner (talk) 20:49, 6 September 2009 (UTC)
Also, FTR, it isn't an abbreviated form. We may know it is derived from kilo binary, but the standard just says kibi (and talking about kilo-binary-bytes would doubtless cause more confusion than already exists).
--Joe Llywelyn Griffith Blakesley talk contrib 00:23, 28 February 2006 (UTC)
I really gotta learn how to read those IPA symbols. For the record, I'm from the midwestern U.S. and I would pronounce it like... well like how you'd expect "kibby" to be pronounced. Kind of like a girl's nickname or something. Vid the Kid - Does this font make me look fat? 16:29, 14 August 2006 (UTC)


If KiBi means kilo binary, shouldn't it stand for 1000 bits?

No, you should understand "kilo binary" as "a binary number that is close to kilo" ;-)
Wouldn't that be a kilobit? WalrusMan118 18:22, 21 March 2007 (UTC)

In SI, kilo is abbreviated to a mere k. Isn't kibi abbreviated to "ki" rather than "Ki"?

Well. I stand corrected from the references on Binary prefix. Shame though.


Regarding the gigabyte vs gibibyte, that is a substantial difference, which quite disappointed a friend of mine once. I have noticed that it seems to be common practice to advertize hard disk capacity in a unit called GB with a footnote defining it as "Billion Bytes". Maybe this should be noted in the article. Personally, I think if you say "gigabyte" or any other -byte unit, the 2n interpretation should be assumed, meaning separate -bibyte units shouldn't be necessary. Vid the Kid - Does this font make me look fat? 16:29, 14 August 2006 (UTC)

It also should be noted that the difference between the binary sizes and the decimal ones becomes larger when the sizes grow. So the problem will worsen. Adding this to the article, with a link to http://en.wikipedia.org/wiki/Binary_prefix#Deviation_between_binary_and_decimal_interpretations. Orafu (talk) 17:05, 29 October 2009 (UTC)

Need for Kibibyte[edit]

I cannot conceive of any reason for special consideration in the computer realm for changing the meaning of the kilo prefix to mean 1,024 versus 1,000. Just because memory comes in quantities of powers of two does NOT mean this special dispensation for marketers is justified. A company I once was employed by sensibly took the decimal number of kilowords (in that context, it was words rather than bytes) and divided by 1,000. Thus 524,288 words was quoted as 524K words. Simple, easy, everybody did it and understood it.--72.75.86.192 08:38, 28 December 2006 (UTC)

Personally, I think this word is silly. The only place where "kilobyte" means anything OTHER than "1024 bytes" (and megabyte meaning 1048576 bytes and so on) is in quoting storage capacity, and (IMO) disk marketers just did that to inflate their numbers. I resist using this unnatural word to overcome a language flaw created by a marketing department to inflate perceived hard disk capacity. 24.6.224.12

When people have to calculate binary sizes, as a programmer or similar, you want to be exact, and it is much easier to refer to powers of two than to have to round the result. If you take four times the size of 524 kilowords, the result is not 2096 kilowords, but 2097 kilowords. Also, in technical papers people tend to be more exact, and it is useful to have an exact unit. Admittedly, in conversations it is sometimes easier to refer to "binary kilobytes" or "power-of-two kilobytes" that to "kibibytes". But the reason often is that the term "kibibytes" is not yet used that often. Also, the notation "kiB" is not perfect, because it does not really remind you of the correct pronounciation, or the meaning of the term. Still, it is the best notation of the proposed ones. Orafu (talk) 16:58, 29 October 2009 (UTC)

Never heard of it[edit]

25 years heavily involved in computers and electronics, widely read, and a degree in Electronic Engineering, and I've never come across this term before. Kilo, Mega, and Giga bytes always referred to the tenth, twentieth, and thirtieth powers of 2, because anything else is impossible to work with when you are using binary addressing. —The preceding unsigned comment was added by 192.118.76.130 (talk) 18:06, 20 January 2007 (UTC).

The reason it was established is largely that manufacuturers of memory and storage realized they could better market their drives in decimal kilobytes (now gigabytes). Kibibytes, gibibytes, etc are entirely unambiguous and can't be fudged. For an example, see the page for the Mac Pro where the fine print says "(1) 1GB = 1 billion bytes" http://store.apple.com/1-800-MY-APPLE/WebObjects/AppleStore.woa/wa/RSLID?mco=2CB5E8C0&nclm=MacPro 142.157.192.86 07:11, 4 February 2007 (UTC)
I have never heard of it either, I think it makes pages such as the page for the Macintosh unnecessarily confusing. Darwin16 16:12, 24 January 2007 (UTC)
The only place I've ever seen "kibibyte"/etc used has been on Wikipedia. It's true that HD manufacturers use powers-of-ten for advertising reasons - ok, why does that mean we have to change (and it is a change) all of our terminology? The two meanings can co-exist, or pressure can be put on them to change. Even memory manufacturers don't play this game. All "kibibyte" does is confuse people. 23 years professional experience here, ~20 years of reading EETimes.
Wikipedia is supposed to document topics, not be a vehicle for changing terminology. This strikes me as a crusade. — jesup 04:29, 14 February 2007 (UTC)
Agreed. Regardless of whether anyone thinks that "kibibytes" are a good idea or not, they are not conventional in any context, anywhere on earth. The convention in deceptive advertising is to refer to thousands of bytes as kilobytes. The convention everywhere else (academically, professionally, journalistically) is to refer to powers of two under those SI prefix names. Of those two, the latter (standardising on professional consensus, rather than on deceptive advertising practices) is clearly the preferable convention on Wikipedia or anywhere else. But regardless, standardising on a term which no one uses, which has been invented to emulate the terminology of deceptive advertising, is ridiculous. This is an extremely silly crusade. --Yst 21:10, 6 March 2007 (UTC)
Silly or not, if you're objecting to the use of 'kibibyte' in Wikipedia articles, this is not the right place to try to change that. BTW, 'kilobyte' may be the convention, but it's also wrong — The International System of Units makes this explicit (see Binary prefix#_note-BIPM). While I appreciate what you're saying, using SI prefixes incorrectly throughout an entire encyclopedia is at least as silly. --StuartBrady (Talk) 21:46, 7 March 2007 (UTC)
This is not the place to change it: true, the place is WT:MOS - but if you try to make the argument there, the editors who make MoS issues (and the correction thereof on Wikipedia) their life will tell you it's all been discussed before and consensus is that you're wrong, and it's not open for further discussion or change because the editors who made the consensus can't deal with the continuous trickle of annoyed editors and users who've run into this coming to complain. Which in a way is true: each (talk) page on WP develops consensus on issues based on the active editors thereof and those who actively monitor the page. The problem is that guideline articles are inhabited by (and consensus is "made" by) those who make said guideline a major aspect of what they do on Wikipedia, while those articles indirectly affect many or all other articles - but the editors of all those affected articles only get involved in guideline pages when someone uses a guideline to override "consensus" of page editors. And I use "override" specifically. There are positives and negatives to this: it avoids "me and my friends like doing everything our way", but it also (in cases like this) leaves a bad taste in the mouth of serious editors. This has cropped up multiple times: this issue here; on WP:EL with youtube links and a small cadre of editors removing almost all links (often including links to videos that are not copyvios), etc — jesup 05:13, 27 March 2007 (UTC)
Opinions clearly differ on this. As in other cases where people may have different approaches and opinions on what is right (e.g. spelling color or colour, or saying a person is English or British), I take the view that changing existing articles is provocative, and revert them. Notinasnaid 22:45, 7 March 2007 (UTC)
I also had never heard of kibibyte until I found somebody had edited my Notepad+ article and was about to revert the vandalism when I saw it referred to the Manual of Style. Another pointless thing which WP is trying to use to rule the world. In the context of that article (referring to a limitation in the original MS Notepad) it is totally irrelevant as the difference in bytes is trivial compared with 'unlimited'. Let's stop this nonsence Dsergeant 06:09, 26 March 2007 (UTC)

But a restriction of 64 decimal kilobytes wouldn't make much sense, would it? The fact that it's a power of two is relevant to the discussion. — Omegatron 00:03, 27 March 2007 (UTC)

Nobody names 1000 bytes a kilobyte. Who and where introduced the term "Kibibyte"? I am programmer since 1991, and I never heard of it. Alone Coder
First of all Yes, they do name 1000 bytes kilobytes. (do we need to cite sources in Talk pages? ;) Second of all, I'm curious, you either figure you know everything, or things don't change. Which is it? Equant 12:20, 14 June 2007 (UTC)
First of all, whom do you mean by 'they'? Second, I've said what I said, nothing more. Third, a small research. Following the only reference in this wiki article, I found some popular article, which gives a link to IEC 60027-2 [1] - address reporting an error. IEC web store does not contain any "IEC 60027-2" string. Try yourself: [2] Alone Coder 16:55, 10 September 2007 (UTC)
I have been programming since 1982, and never heard of a Kibibyte; until today. That's why I came to Wikipedia. The Ubuntu bug report system uses the abbreviation "KiB" and I wanted to know what it was. Those of you saying this page shouldn't exist are wrong. Just because this convention (and resulting article) exists, doesn't mean you have to use it or like it. Equant 12:20, 14 June 2007 (UTC)
The whole kibi thing is nonsensical, confusing, and possibly immoral. Is Microsoft behind this?--Gmuir 23:40, 7 August 2007 (UTC)
Agreed, I move for deletion on grounds of inconsequentiality. Zaphodb777 (talk) 12:03, 21 March 2008 (UTC)
Ubuntu uses both KiB and MiB extensively; look at your System Monitor. I'm not sure if this is the thing with all Linux distributions or not, though I'm quite curious to get some feedback if anyone is using a different distro. More on the topic, these haven't hit mainstream yet, but it looks as though they're here to stay. Just look at IEEE 1541. If anyone can find what the ISO has said about this, please let me know. Crmadsen (talk) 05:02, 1 July 2008 (UTC)
HEY GUYS! Remember what the article explicitly mentions about who defined the damn unit: "...established by The International Electrotechnical Commission in 2000." Take a look at this: When is a kilobyte a kibibyte? And an MB an MiB? In my personal view, the traditional KB = 1024 bytes, etc. customary definition, even if not quite correct, was clear, unambiguous, and useful enough. Hard disk manufacturers should have been forbidden to corrupt the terminology as they did, using deceitful tactics. Regards, --AVM (talk) 22:33, 26 September 2008 (UTC)

Toyabyte[edit]

Why toyabyte is named the same in both conventions if it has the higher difference ? (10^27 vs 2^90) --Kewlito 14:29, 9 February 2007 (UTC)

Weasel byte?[edit]

Who has heard of a "weasel byte"? Only 4 Google hits, none of which are clearly about what is described here. Selling 1000 bytes as one KB is the weaselly thing, and 1024 bytes is fine, so if anything this is also the wrong way around.  --LambiamTalk 07:28, 9 March 2007 (UTC)

Since I wrote the above, the section this referred to has been deleted.  --LambiamTalk 08:51, 9 March 2007 (UTC)

Linguistically[edit]

Language is a convention. Words mean what native speakers actually use and understand them to mean. So far as I can tell from my usage and from the usage of other native speakers, "kilobyte" is perceived to mean 1024 bytes, and there are good reasons for this usage.

  1. It more closely reflects how computer storage is actually allocated and used.
  2. It is also how every computer I have every used reports storage capacity and availability.
  3. Since the vast majority of users see storage reported on their computers much more often than they buy storage, it is the meaning they encounter most often.
  4. The only time I ever see "kilobyte" used to mean 1000 bytes is in marketing, and even then RAM is still measured in powers of 2, because measuring it in powers would be too awkward, requiring the use of fractional amounts in something that is perceived as a count noun.

It would be less confusing for users if storage makers used terms like "20 billion bytes" instead of "20 GB." For my own purposes, I use "marketing kilobytes" or mKB for 1000 bytes, though I'm sure this would never catch on. Bostoner (talk) 20:49, 6 September 2009 (UTC)

Merge[edit]

Can we centralize this discussion somewhere for all the articles? — Omegatron 17:45, 10 May 2007 (UTC)


Merging this would be silly at this time 169.252.4.21 08:30, 29 March 2007 (UTC)


The article is good at it is. A merging would cause more confusion. ~Anonymous 24, April 2007

These unit articles have very little content in them that isn't already in other articles. They should all be merged into something. Maybe Units of computer storage or Units of data? — Omegatron 21:57, 24 April 2007 (UTC)
I observe the merge was done today without consensus. I undid it. One key reason is that many articles link to the unit name in the context of using it. The term is certainly unfamiliar to many, and this article helps by providing a focussed discussion. Notinasnaid 11:13, 7 May 2007 (UTC)
Yeah. If the merge is going to be done it needs to be done right. For one, each unit should have a section or separate definition, and the individual articles should redirect to each definition using anchor tags, so that people can link like KiB and go right to the definition. — Omegatron 17:33, 10 May 2007 (UTC)
  • Binary prefix contains all the information in this article, and then some. It's not really useful to have fifteen articles on the same topic. Note how this article is essentially similar to mebibyte and gibibyte. This article is not a focused discussion, and causes confusion through divergence from other articles containing the same information. Wikipedia has redirects for a reason. It is perfectly reasonable for a link "kibibyte" to point to a page explaining such kibibytes, without requiring that that is the title of the page. >Radiant< 12:11, 7 May 2007 (UTC)
    You wrote "This article is not a focused discussion, and causes confusion through divergence from other articles containing the same information." What is the specific confusion that it causes? Notinasnaid 17:39, 10 May 2007 (UTC)
    For myself, I don't think it causes confusion, but it duplicates the same content needlessly eight times and allows the articles to evolve in an inconsistent manner. The definitions and abbreviations for each unit only take up a few lines. The rest of the articles are just background information, which would be better presented in a single place. — Omegatron 17:45, 10 May 2007 (UTC)
  • I think that KB and KiB deserves their own article because they are confusing binary prefixes. They are part of a scientific system that strive to avoid confusion and instead brought ambiguity. So I think that they deserve a particular article, instead of concentrating the discussion in the binary prefix article. Regards Loudenvier 16:45, 10 May 2007 (UTC)
    Their definitions would be exactly the same, just combined into a single article that explains the issue all in one place. And, for the record, the ambiguity has always been there. :-) — Omegatron 17:33, 10 May 2007 (UTC)
    I have to make a clarification: KiB has no ambiguity, only KB is ambiguous. It would be nice to have a separate article about the ambiguity of KB (and MB etc) and the introduction of the "i" prefixes (KiB, etc). Perhaps a separate article for each prefix is really redundant, although an article about the reasoning behind the new "i" prefixes is not. Loudenvier 17:11, 11 May 2007 (UTC)
    KiB has no ambiguity, only KB is ambiguous.
    Right.
    Wrong. KB has no ambiguity. 1 KB = 1024 bytes, 1 MB = 1048576 bytes, always. The perceived ambiguity arises only because some, largely manufacturers of storage devices, have a reason to use the terms deceptively. Lies are lies and should be recognized as such -- they do not cause ambiguity in the language itself. Ramorum (talk) 20:26, 3 February 2008 (UTC)
    It would be nice to have a separate article about the ambiguity of KB (and MB etc) and the introduction of the "i" prefixes (KiB, etc).
    an article about the reasoning behind the new "i" prefixes
    It's already covered in Binary prefix.
    Perhaps a separate article for each prefix is really redundant
    Exactly.  :-) There is a significant amount of content specific to each unit, like the three different definitions of megabyte, but I don't think that means we should reproduce all the background info for each separate unit's page. That unique content should be a section in a larger article instead. — Omegatron 03:11, 12 May 2007 (UTC)

Being introduced to this topic, I read both the kibi and kibibyte pages. The latter first. It cleared up all my misunderstandings. The kibi page is an overwhelming mass of information, where this page gives a fly-over view which explains the concept before being thrown into the deep end. My vote is not to merge these two pages. -- onno@itmaze.com.au, 12 June, 2007

Since this page must necessarily describe KiB in terms of MB and GB (and they must make reference to this page) it would make sense to merge them all into one page, with sections as necessary for the above issues. -Gomm 18:37, 12 July 2007 (UTC)
I agree that it would make more sense to centralize this information. At its core, all you need to do is display the table showing the individual unit names and their values in each category, and then have explanatory text about the general derivation of the words, common misuses, differences between Kebibyte and Kilobyte, etc. Do we have individual articles for "gram", "kilogram", "centigram", "milligram", etc? I doubt it - I don't see why units of data storage should be treated differently than units of length, weight or speed. — KieferSkunk (talk) — 20:40, 15 August 2007 (UTC)
This is a bit outdated. Now we have article on all SI units: kilogram, kilometer, nanometre, etc. We have stubs for rare units such as 1 zettametre. --Enric Naval (talk) 13:32, 21 August 2014 (UTC)

A new standard[edit]

This is not yet a standard, but it removes the perceived ambiguity and gives people something to say if they want a unit for 1000 bytes and it is thus recommended for those who must refer to powers-of-ten quantities of bytes.

103 bytes = 1000 bytes = 1 ψKB (or 1 pKB or 1 dKB) 106 bytes = 1000 ψKB = 1 ψMB

ψ- or p- for pseudo, portraying the fact that 1000 bytes is not 1 KB but is close to 1 KB, hence one pseudo-kilobyte. Alternatively, d- for decimal, portraying the fact that 1 dKB = 103 bytes

This new terminology would be consistent with the already-existing standard that 1 KB = 1024 bytes, 1 MB = 1024 KB, etc.

Ramorum (talk) 20:37, 3 February 2008 (UTC)

I believe dKB was proposed too, but the KiB proposal was the one adopted by the standards organizations. ψKB is the sort of thing you publish on your own site or blog. Wikipedia is not the place to propose new standards. — Omegatron 00:17, 22 March 2008 (UTC)

Where to discuss use of this word in wikipedia[edit]

Where can I join (what I assume must be an active) debate about whether new wikipedia pages should use kilobytes, which are in active use, or kibibytes, a term derived from a form of cat food? I mean Wikipedia is the only site I know of which uses these words, the only other places I've seen them mentioned (tech forums etc) are where their lack of uptake has been ridiculed. Please end this crusade already. Btw, that was a serious question. It's an issue that does bother me, and I would like to read the formal discussion that led to so many wikipedia pages using these abominable terms. Themania (talk) 15:17, 6 March 2008 (UTC)

If only all questions were this simple to answer. The presently accepted guidelines are at WP:MOSNUM, and these are frequently discussed on the corresponding talk page. You will find the debate is heavily polarised. Please feel free to dive in with your views. Thunderbird2 (talk) 15:33, 6 March 2008 (UTC)
You don't seem to be interested in discussing at all. I'd recommend that you read the existing discussions and the available information, before you join any of them yourself. --217.87.75.67 (talk) 14:49, 28 March 2008 (UTC)
If you read to the end of his paragraph, that does seem to be his intent. 12.216.36.141 (talk) 14:08, 12 April 2008 (UTC)

Incorrect name?[edit]

The proposed "kibibyte" which stands for "Kilo Binary Bytes" is a misleading term. What can this stand for? Possible interpretations:
1. Kilo (Binary Bytes) - a thousand of binary bytes. This is obviously incorrect.
2. (Kilo Binary) Bytes - "Kilo Binary" stands for a "thousand of binaries". The result is "2000 bytes". Incorrect again.

The correct approach would be to prefix the "Kilo" with the "Binary" modifier. So the term should be "Binary Kilo Bytes".

Please consider this. Current term is strictly speaking unacceptable.

Thank you. —Preceding unsigned comment added by 193.238.96.4 (talk) 09:15, 28 January 2010 (UTC)

The correct interpretation is that kibibyte does not stand for 'kilo binary byte', but that the prefix 'kibi' was constructed as a mnemonic from the words kilo and binary. Kibibyte then is a composite of the now standard term 'kibi' treated as a mathematical entity just like 'kilo' in the SI system. To explain its origins, these prefixes are still often explained by referring to 'kilobinary', etc., to help memorizing them, and even NIST does that, but to decompose the unit kibibyte into the original components and to try to analyze the meaning in turn is wrong. We don't revert to the origin of 'kilo' in kilobyte either to explain it. Kbrose (talk) 05:19, 29 January 2010 (UTC)

Yes, I see. That makes sense. Thank you. —Preceding unsigned comment added by 193.238.96.4 (talk) 16:29, 31 January 2010 (UTC)

Another infobox fail...[edit]

Kilobyte is 1024 bytes or 2^10. a value of 2^10 conveys meaning only in binary. How about:

Kilobyte 10^3 Kibibyte 2^10 Decimal approximation: 1000 ( 2.4% error ) and how about Capitalization? The first kilo is in lower case?

I will be back to make this CLEAR 69.232.205.64 (talk) 05:14, 2 March 2011 (UTC)

Problem with Facebook Ripoff[edit]

On facebook the 2^10 is rendered as 210. What a shame. Who could be contacted?

Janburse (talk) 11:42, 15 June 2011 (UTC)

The article is currently a nonsense[edit]

A kibibyte is currently accepted in the computing world as 1,024 bytes regardless of who may like to believe otherwise (including the IEC). Disc manufacturers may prefer to use it in its more technically correct form as 1000 bytes for marketing reasons, but that does not change the accepted usage (disc drive manufacturers have to specifically state their non standard usage in most jurisdictions). The use of 'kibi' for the former and 'kilo' for the later is not much more than a proposal at the present time. Whether kibibyte is gaining more acceptance is moot as it certainly is not universally accepted as a valid unit and in reality is rarely used. Wikipedia policy (at WP:MOSNUM recognises this and discourages its use precisely because it is not generally accepted. The article is currently written as though it is universally accepted. 86.145.240.230 (talk) 12:12, 18 October 2013 (UTC)

Naw, I figure they'll state it again and again until blue in the face, take a breather, then keep saying it some more until everyone believes it. This is about like trying to change the convention so that electrons are positive and protons negative, to make electricity make more sense. Sorry guys, an arbitrary decision was made when producing the foundation of the field and now is a bit too late to change it. Imagine if red lights meant go and green lights meant stop in a specific city with no warning to tourists.  :D 2601:1:9280:155:45FF:30A1:76F8:2112 (talk) 21:03, 8 June 2014 (UTC)
It's starting to become more common. Certain programs now reference KiB, MiB, etc. in both output and input options. I have always had to mentally differentiate between kilo=1024 and kilo=1000 based on context. Any computer professional has done this their entire career. For us its hard to understand why the controversy.

24.163.123.33 (talk) 20:07, 18 August 2014 (UTC)