Archive 5Archive 7Archive 8Archive 9Archive 10Archive 11Archive 15

Hard drives "always used decimal"?

Following my previous contribution (now in Archive 1), it seems that in modern times, all hard disk manufacturers use, as far as possible, binary GB. This has not always been the case. Take, for example, Maxtor's D540X 4D. The "80 GB" model has 158816*16*63*512 = 81964302336 bytes. The product manual instead gives the capacity as "82.0 GB" (it also, interestingly, says 30735581184 bytes is "30.0 GB").

Then there are older drives, the 90431U1 model which is 4343463936 bytes, or about 4.04 GiB (funnily, the "173" model is about 16.21 GiB, and 17,002,440 KiB. The Quantum Fireball 4.3AT CR43A101 is 8895*15*63*512 bytes (about 4.304 decimal GB, 4.008 GiB).

While I haven't found a manufacturer that defines KB, MB, or GB as anything other than powers of 1000 (or something like "one billion"), manufacturers have, for whatever reason, made hard drives which are a little over an integer multiple of 106210 or even 230 bytes (I don't have an example of 103220 handy, but I'm sure it's out there). They knew that 1024 was important to some people. Perhaps this is because it's far more convenient to go between KiB and block counts than not-actually-SI KB. It's also pretty obvious to me that Quantum wanted 4 binary GB, but also to present the drive size in marketing GB - the marketing reason is also why hard drives are just a little over a multiple of 1010 bytes, while just a little under a multiple of $10.

And then, the use of K, etc, as powers of 1024 is not "incorrect". It is not SI. It never claimed to be SI. It's confusing, sure, that 1K = 1024, but it's no more "wrong" than using m to mean mile. Meaning is based on context, and the context has been largely binary ever since we stopped using BCD.

I would change these, but the article's HUGE. Elektron 22:23, 22 August 2007 (UTC)


it seems that in modern times, all hard disk manufacturers use, as far as possible, binary GB

They do?

it's no more "wrong" than using m to mean mile.

It's no more "wrong" than using "kg" to mean "437 million graham crackers". I can define it to mean whatever I want. As long as I don't claim it to be SI, it will not be "wrong" ... right? — Omegatron 23:30, 22 August 2007 (UTC)

In modern times, the available capacity is a production yield decision. All drives have somewhat more capacity than specified; they are set to provide a particular value in the manufacturing process and if u knew the "secret" commands you could gain access to the additional capacity. IMO, the set capacity is placed a little above the stated capacity in 1010th units to allow for the loss of capacity due do defects grown in use. Therefore, over the life of the drive, the customer is never denied the advertised 1010th capacity. We would have to talk to manufacturing folks to see exactly what they are trying to achieve but I doubt if it has anything to do with binary units. Tom94022 00:09, 23 August 2007 (UTC)

I meant that HD manufacturers use 109 GB. The addressable sectors (and what I refer to) never include spare sectors - they're transparently remapped on a write to a bad block, and presumably on a read from a block with a consistently high error rate. You never see them. Therefore, over the life of the drive, the customer is never denied the advertised 1010th capacity. does not follow (and I'm not sure what you mean by "1010th") - the number of spare sectors is pretty small, and I suspect less than 1%. What I'm talking about is how hard disk manufacturers have, quite consistently, made drives in near-multiples of 1024 or 1024³, instead of the "usual" 1000³, and how they've sometimes advertised a 81.96 109 drive as "80 GB", when they could've called it 81 GB (and they call it 82 in the manual) or just not added the extra sectors. Yes, they say "1 GB = 1000000000 bytes", but they give you a little extra for no apparent reason. ⇌Elektron 13:53, 5 September 2007 (UTC)

I have replaced the bare links used as references in this article with proper citations per Wikipedia:Footnote. The essential problem with bare links (a URL with no other identifying information) is that they frequently break, and without data like news article titles or website page titles, it is often impossible even to know where to look for a replacement source. (Archive sites like the Wayback Machine or Google sometimes help, but often do not.)

For example, the following statement has become unsourced because its old ABC News link no longer exists, nor appears to be readily accessible in an archive:

Western Digital offered to compensate customers with a free download of backup and recovery software valued at US$30. They also paid $500,000 in fees and expenses to San Francisco lawyers Adam Gutride and Seth Safier, who filed the suit.

The editor contributing this "source" failed to provide any information that would allow us to find a replacement for it. (It's entirely possible that one of the existing sources has this information, but I leave this to someone prepared to do a thorough source review to fix. I'm just making it easier for such a review by properly citing the sources.)

In short, bare links should be never be used as sources for Wikipedia articles. You don't necessarily have to create a fully filled-out citation, but at least include basic title and/or descriptive information (date, author) with the reference. ~ Jeff Q (talk) 16:47, 9 September 2007 (UTC)

Alternate proposal

Should we mention the "dikilo"/"dimega" proposal in this paper? http://www.cl.cam.ac.uk/~mgk25/information-units.txtOmegatron 00:55, 11 September 2007 (UTC)

As much as I like Markus Kuhn, I don't think it satisfies any requirements for notability (it's a paper published on a website and not endorsed by anyone else). It also breaks down at 1000^30, where the "next power of two" is 1024^29.9, not 1024^30. It also doesn't solve anything that the IEC proposals don't, and could cause more confusion — "b" almost universally means bit, and if you want to use a "SI" Mbyte, there's no way of saying "M means 10^6" except by spelling it out. The byte's a de-facto unit anyway, and converting between advertised Mbps to "real" 10^6 B/s is often a factor of 8, 10, ~20 (WiFi), ~16 (USB 2.0), or something equally strange. ⇌Elektron 11:30, 11 September 2007 (UTC)

War has never been so much fun

My edit has just been reject as "vandalism". This is ridiculous! We can disagree all day whether M means 1000^2 or 1024^2 but this sentence is a quite different issue. In any case, 'M' is definitely not a binary prefix. The binary prefixes are Ki, Mi, Gi, Ti, Pi, Ei. Therefore, the sentence "Binary prefixes are often written and pronounced identically to the SI prefixes, despite the resulting ambiguity." is not only incorrect, it makes no sense at all. Who pronounces MiB as MB? I pronounce the former mib as in MiG (or meh-bee-bite) and the latter em-bee (or mee-gah-bite). I don't see how you could pronounce any of them identically. You could add an introduction like "SI prefixes are still frequently incorrectly used for values that are powers of 1024 instead of the unambiguous binary prefixes." but that's later explained in the article, thus redundant. --217.87.98.171 00:19, 1 November 2007 (UTC)

Your edits have been reverted as vandalism because your IP and your other "anonymous" IP:
  • Have been used as for single purpose edits against MOSNUM policy.
  • Have been asked by at least three different editors to not make changes to contrary MOSNUM policy yet you continued to do so.
  • Have been used to write personal attacks against other editors on their talk pages and in article talk pages.
  • Have been asked not to harass those editors on their talk pages yet you continued to do so.

Fnagaton 00:42, 1 November 2007 (UTC)

Even if you consider some of my edits as vandalism, that doesn't justify reverting whatever I edit. There is nothing wrong with "single purpose edits" and it's only against WP:MOSNUM by your definition. Also you have to specify the exact revision because WP:MOSNUM has been modified not so long ago and will likely change again. Furthermore, you've shown hostile behaviour towards me by talking in bad faith about my edits on the talk page of your friend. You are accusing me of being a sock puppet just because I do care about IEC standard prefixes and have the same opinion as someone else on this topic when there are in fact millions of people and at least dozens if not hundreds of wikipedia authors sharing the same opinion. You are even accusing me of stalking now. Stalking is a criminal offence! In fact you're doing exactly the same, you're monitoring my edits. I've told you more than once that I am not User:Sarenne and not in anyway related to this user. I do however assume that this user was banned in bad faith and if this user misbehaved it was very likely due to provocation by your group and zero-tolerance regarding a flawed and recently modified policy WP:MOSNUM. The policy is flawed in so far that use of IEC standard prefixes is not a question of style but correctness and adhering to current standards. The policy is also badly worded which allows you to get away with your constant reverts. "There is consensus that editors should not change prefixes from one style to the other, especially if there is uncertainty as to which term is appropriate within the context". As said it's not about "style" if you write 64 kB instead of 64 KiB. It's about style if you write 65.5 kB instead. The word especially makes no sense here. If an author is not sure, he certainly should not edit it at all before finding sources to make things clear. Then it says "When this is certain, the use of parentheses for IEC binary prefixes, for example, "256 KB (KiB)" is acceptable. Why are they merely acceptable? Obviously, the current wording puts anyone who likes to use the modern prefixes for clarity and correctness into a big disadvantage despite being an international standard. It is not up to Wikipedia or some contributers to decide what is standard and what isn't. How can you even argue that nobody knows or uses this standard (which is wrong) if Wikipedia which is supposed to educate people and provide objective and up-to-date knowledge, is now strongly opposing spreading this knowledge and sticking to outdated conventions? --217.87.98.171 01:25, 1 November 2007 (UTC)
And once again Sarenne, this is not the place to push your views and agenda on kibi/mibi/gibi useage at Wikipedia. Same old stance, same old tactics, same old attitude, etc. As you were explained by Fnagton on your other ip related talk page, these things are set by WP:MOSNUM, which is the standard here. Take it up there. Your arguments, right down to the very wording, have all been heard before - all by Sarenne. You're not saying anything new, or doing anything different than the last time. And there's a lot more involved then just a "few people" or wording that was slanted in a certain favor. The wording was reached through consensus. As you were also told, "MOSNUM has been changed due to debate and consensus being formed by a large group of editors agreeing what should be done. You are also mistaken, the pro-IEC binary prefix arguments do not hold water and that is reflected in the result of the debate and consensus." To borrow some of your usual sarcasm: Any further questions? Then take them to Wikipedia talk:Manual of Style (dates_and_numbers) --Marty Goldberg 07:42, 1 November 2007 (UTC)
Look, I'm not insulting you the least when I say you are paranoid. How often do I have to tell you guys, that I have absolutely nothing to do with Sarenne - except that we apparently share the same view on the binary prefix issue. Further, I am not insulting you but just stating facts, that you do not even understand basic English sentences. My native tongue is German. What does this tell about you? Nothing you could be proud of. Wikipedia policies are no standards by any means. They are guide lines and can, in fact, be modified - even inversed - any time, any day. Duck and cover! Yes, there's actually nothing new to say about binary prefixes. Everything has been said but you're still not accepting the facts. You're even trying to add confusion where there was none by arbitrarily mixing MB and MiB, so that any lay-man will never get the difference and consider them synonyms. You are not trying to spread knowledge and get rid of myths, no you are trying to reinforce these myths and hide knowledge. There are even such non-arguments as "because nobody knows what MiB means, it adds confusion". What utter nonsense! It is linked directly. Do you think readers are so stupid that they get confused by a single term they don't know? Then what's the point of Wikipedia at all if you only dare to expose people to facts (or non-facts) they already know (or assume to know)? "the pro-IEC binary prefix arguments do not hold water" Nice phrase but pure bullshit. The meaning of MiB is unambiguous for laymen and "experts". That alone makes it extremely useful. Furthermore, it is not up to you or anyone else at Wikipedia to decide what is standard and what isn't. How often do I have to repeat this simple fact? "and that is reflected in the result of the debate" You confuse quantity with good argumentation. "and consensus" Even WP:MOSNUM says that there is no consensus. Or did you change it during the last 12 hours? By the way, en.wikipedia.org is only a part of Wikipedia. If you actually checked one of the other languages, you'd see that in some of them the IEC prefixes are accepted and in wide use. But I guess explaining why English men have a problem with SI units would be considered anti-american/british but maybe you get the idea: Euro UK USA Inferiority complex --08:46, 1 November 2007 (UTC) —Preceding unsigned comment added by 217.87.98.171 (talk)
The current consensus in WP:MOSNUM#Binary prefixes is that "There is no consensus to use the newer IEC-recommended prefixes" and also "There is consensus that editors should not change prefixes from one style to the other" and also "stay with established usage in the article, and follow the lead of the first major contributor.". This also has parity with the other related MOSNUM guideline which says "use the units employed in the current scientific literature on that topic.". The meaning of this guideline is clear that the terms used in the majoirty of relevant reliable sources is to be used. In all the articles you and your other "anonymous" IPs have edited this means you must use KB, MB, GB, kilobyte, megabyte, gigabyte as the primary units used in the article. If you don't agree with the policy then stop your "anonymous" IP hopping, stop your threats to circumvent the system, create a user name and debate the topic on WT:MOSNUM. What you propse increases the ambiguity in articles by using terms that are not widely used. Fnagaton 10:26, 1 November 2007 (UTC)
I am not hopping. Maybe you're not aware of the fact, that in Europe most consumer internet accounts have a dynamically assigned IP address. I've also explained that this is not a style issue, it's a question of correctness. The meaning of kB, MB, GB is not clear at all because you're ignoring the fact that many literature features footnotes or explanations regarding these units because use of the prefixes does not conform to use of them in any other field of science. It has also been explained that "obvious to an author" is not equivalent to "obvious to a reader". Please, see WP:OBVIOUS. Also, please, look up the meaning of ambiguity because it is not the same as ignorance as you seem to imply. I will try to fix WP:MOSNUM at a later time but not by now because it requires to inform a lot of people about the discussion and encourage them to get active because otherwise your team will simply overpower facts with quantity as last time. Last but not least, I'd like you to stop spreading lies about my identity, my intentions or edits on the talk pages of your friends. --NotSarenne 18:40, 1 November 2007 (UTC)
By your own admission "You're not going to ban me this way because it's child's play to circumvent such a silly ban.". You write comments like Sarenne, you edit like Sarenne, you use the same ISP as Sarenne and you attack the same editors who got Sarenne banned. KB, MB, GB, kilobyte, megabyte, gigabyte as used in those articles are correct. Their use matches that used by relevant reliable sources. By the way you are not allowed to WP:CANVASS which "is sending messages to multiple Wikipedians with the intent to inform them about a community discussion." Fnagaton 19:11, 1 November 2007 (UTC)
You know very well that you threatened to ban me first which is a policy violation as well. So we're even in this regard. There are a few million people using the same ISP as me. If Sarenne has the same native tongue as me, it's not surprising if my English style is similar. Like-wise, there are many people supporting the IEC binary prefixes, for example, the IEC members. I don't accuse of being a sock puppet of specific users albeit they all show the same hostile attitude and disagreement with my position. I have not violated WP:CANVASS or claimed to have any such plans. If you actually read it, you'll see that your claim is plain wrong and this isn't the first time, you have waved a policy in front of me inverting the meaning of the actual policy because quite obviously you either didn't read or at least did not understand it. --NotSarenne 20:41, 1 November 2007 (UTC)
Actually no, you attacked editors who reverted you when following policy by accusing them of using sock puppets, and that is a policy violation. You vandalised my talk page with your attack. Your editing record demonstrates you are at fault here. I have told you before to not harass me and other editors yet you continue to do so. You will stop now. You admitted to having plans to canvass at "18:40, 1 November 2007" above when you wrote "it requires to inform a lot of people about the discussion and encourage them to get active ". Fnagaton 21:08, 1 November 2007 (UTC)
That my very well be your view of reality but I can assure you in my reality, you used the term "sock puppet" first and then I used it in a reflective manner. If you are allowed to accuse me of something, I'm certainly free to do the same. I do not actively accuse you or your friends of using "sock puppets" though. It's just every time you throw dirt at me, I'll throw it right back to you. Fair, isn't it? So far you have informed a lot more people than I have and you used a huge amount of prejudice. Do you even realize that you sound like a broken record? I came here to improve the article within the tight bounds of the current WP:MOSNUM policy but you're constantly turning this into a feud. Apparently your vendetta against Sarenne and IEC prefixes has made you deaf for any kind of facts and arguments that do not fit your personal preferences. --NotSarenne 00:37, 2 November 2007 (UTC)
You are completely wrong because you used the term sockpuppet first. Your use of the word combined with mentioning old binary prefix history, something Sarenne would know and a new user would not know, is the final straw that made me conclude that you are a sock puppet of Sarenne. Here is the proof, firstly in chronological order all my edits that I addressed to you and your previous "anonymous" IP edits: Revert changes [1] [2] [3] MOSNUM explanation [4] [5] [6] [7] [8] You revert someone elses change [9] You first mention "sock puppet accounts" here [10] Also in this edit you mention the "anti-kib tag team" and "how few people contributed to it and the weak arguments" these phrases plus your previous edits and comments give very strong evidence that you are Sarenne. The final straw was your personal attack later on in that comment, you are a sock puppet of Sarenne. So now I have proven you to be wrong I demand you retract and apologise for your lies that you are writing about me. What you are doing is deliberately misrepresenting me and as such you are acting against policy, so stop it, retract all your lies (which would be all of your recent edits) and then stop responding to my edits because you are harassing me and other editors. Fnagaton 15:58, 2 November 2007 (UTC)
It seems you're right that I used actual term sock puppet first. This was however in response to your comment here: [11] which implies that you consider me a potential sock puppet. I did then check some of your recent edits on the topic, about User:Sarenne and discussions about the binary prefix issues You and others have been accusing others of sock puppetry quite a lot there in cases that were similar to my edits. I've noticed with these few checks that you and others are frequently getting emotional and rude concerning the IEC standard prefixes, So it was just a question of time before you'd use this term against me. Do you seriously deny that you are anti-kib - short for against the IEC standard binary prefixes - and are teaming up with a couple of editors to prevent use of them? You should also read Talk:MB and realize that my edits to this talk page were alright in accordance to the guidelines. Further, I clearly have not violated any policies to the best of my knowledge under this account. Every single time that you and your friends accused me being in violation of a policy, you and them were wrong. I will certainly not apologize for anything, even if I made some minor mistakes, especially due to your extremely rude and childish stance against me. I believe at some point we were even in concern to rude behaviour but at this time I consider me the victim of your harassment, if you keep insisting that there is any harassment going on. --NotSarenne 17:17, 2 November 2007 (UTC)
Do not mistake "rude" with "being efficiently accurate with my arguments". Do not forget you used multiple personal attacks first. Stop with the "implies" rubbish and trying to misrepresent what I write with your ad hominem. The facts prove you are wrong so apologise and retract your edits. Your continued stalking and harassment only go to prove how bad your behaviour is and how I am justified for creating a sock puppet report about you. Your refusal to apologise for your actions shows this. I did accuse Sarenne of using a sock puppet, by adding facts and references to his sock puppet report, and the result of that was the user was banned. Fnagaton 17:30, 2 November 2007 (UTC)
Rubbish you say? I wrote "I wonder how many sock puppets you have." That is not the same as a claim and it's roughly the same as implying that I am Sarenne by writing "I hope it's not Sarenne coming back" and it's none of your business because I didn't write that to you but Marty Goldberg. Oh wait... just kidding. Anyway, regarding Wikipedia:Sock_puppetry, I see that you're again reading a policy to the letter but don't get the idea. I knew the term since long ago and I have occasionally been editing articles, reverting vandalism etc. without using an account for years without causing any conflicts. Perhaps because I rarely edited any controversial article or just didn't check what happened to my edit. So I'm not unfamiliar with Wikipedia. You're interpretation is that this is "another prove" that I must be Sarenne. Oh well, believe whatever you want, but please, cease your vendetta. You're smarter than me, aren't you? Shouldn't you give in then and stop wasting your time? --NotSarenne 18:19, 3 November 2007 (UTC)
The fact is you made direct accusations which are untrue without any evidence. You have been proven to be wrong and you still fail to retract and apologise. You are acting against policy. The rest of your edit misrepresents that actual facts of the situation. Fnagaton 18:50, 3 November 2007 (UTC)
Do you confuse talking and stalking by any chance? Would you mind to explain these oddities: Special:Contributions/QuinellaAlethea, Special:Contributions/HyperColony? --NotSarenne 18:30, 2 November 2007 (UTC)
You are not just talking. As demonstrated above you write lies about me and then refuse to retract and apologise when you are proven to be wrong. You personally attack me and other editors and because of your bad behaviour I have told you to stop writing about me and stop responding to me yet you continue to do so. So stalking as it applies to your actions is "to follow or observe a person persistently, especially out of obsession or derangement". Prove you are not stalking me by complying with the warnings I gave you and your other anonymous incarnation. Do not forget it is your actions and bad behaviour that cause you to be warned in such a way. Obviously those others editors are examples of people who disagree with your behaviour. Fnagaton 12:21, 3 November 2007 (UTC)
You should rather link to the history of my talk page, so that readers get the big picture. What the heck makes you believe you can address me on my very own page and then forbid me to write a response on the very same page. It is violating exactly the policies that you were waving in front of me. Further, what do you think gives you the right to forbid me to answer you on whatever page you are talking to me? If you had any kind of honour and actually believed in what you said, you would have addressed those pseudonymous users and asked them to not participate in something that is none of their business. --NotSarenne 17:54, 3 November 2007 (UTC)
The "big picture" as you put it is actually you writing lies about me, there is policy which states you are not allowed to use your talk page to wilfully misrepresent other editors and you are violating that policy. I do have honour and those editors obviously see you are at fault and since they are not attacking me (like you do) and are addressing the fact you are being disruptive then they can comment as they see fit. The fact is you are violating policy and your reply by your own admission also proves that you are stalking me. Fnagaton 18:27, 3 November 2007 (UTC)

About prefixes and units

I'd like to explain why this revert is unjustified: [12]. MB is a unit but not a SI unit because bit and byte are no SI units. k, M, G, T are SI prefixes. They are prefixes because you put them in front of units. A unit itself can be considered a suffix because you put it behind a value. MB is a (scaled) unit but never a prefix. Therefore, the sentence made no sense and I modified it. Maybe sentences don't need to make sense here, so the revert might be okay. Who knows? --217.87.98.171 04:38, 1 November 2007 (UTC)

Once again, that's misreading the si prefix article. Specifically, its talking about the K, B, and G prefix added to the unit (bit, byte, etc.). It is not stating to use K, B, and G alone. The traditional usage is KB (or kB), MB, and GB. --Marty Goldberg 04:48, 1 November 2007 (UTC)
I'm not misreading. The current sentence reads the decimal meanings of KB, MB, and GB are often referred to as SI prefixes. So MB is referred to as a SI prefix? No, it's not. M is referred to as SI prefix. MB is a unit, not a prefix. --217.87.98.171 05:22, 1 November 2007 (UTC)

Enlightened by Cho v. Seagate Technology (US) Holdings, Inc.

I see:

"Seagate has denied and continues to deny each and all of plaintiff's claims, and denies that anyone has been harmed or deserves compensation. The Court has not made a decision on the merits. ..." -- http://www.harddrive-settlement.com/notice-email.htm

"A hearing has been scheduled for Feb. 7, 2008, in San Francisco Superior Court to approve the settlement" -- http://www.computerworld.com/action/article.do?command=viewArticleBasic&taxonomyName=storage&articleId=9045141&taxonomyId=19&intsrc=kc_top

"One gigabyte, or GB, equals one billion bytes when referring to hard drive capacity. Accessible capacity may vary depending on operating environment and formatting. Quantitative usage examples for various applications are for illustrative purposes. Actual quantities will vary based on various factors, including file size, file format, features and application software." -- http://www.seagate.com/content/pdf/datasheet/disc/ds_portable.pdf

Q1. Has Seagate here now dreamed up brief clear neutral-point-of-view that we could/ should copy/ paraphrase into Gigabyte, Megabyte, etc.?

Q2. Should Binary_prefix#Legal_disputes reference this dispute also?

-- Pelavarre 13:28, 3 November 2007 (UTC)

You might want to add that all these lawsuits took place in the USA. This confirms two things: First, the US law system is very abuse-friendly and second, the universal well-defined meanings of kilo/mega/giga are widely-unknown due to refusal of the USA to use SI units. Quite obviously, the wrong companies were sued. Of course, Microsoft and Apple (albeit to a much lesser extent) are so powerful that nobody who's right in his mind would dare to sue them. --NotSarenne 15:09, 3 November 2007 (UTC)

Giga vs. Mega

Gigabyte and Megabyte describe this 1024 = 1000+-2.4% dispute of people gibbering of maybe bytes differently. That difference is confusing and pointless?

At Megabyte, we say 'the term "megabyte" is ambiguous'. At gigabyte, we say "the usage of the word "gigabyte" is ambiguous, depending on the context', and then our two articles go on to diverge farther from there. Admittedly, the audience who gives adequate time & attention will eventually see that megabyte links to gigabyte which links here, but should we be requiring so much time & attention for the reader to walkaway with a correct understanding?

-- Pelavarre 13:28, 3 November 2007 (UTC)

People are free to make fun of things but pronouncing mebibyte as "maybe bite" is just as childish as saying "I pee a dress". It mebi funny but it's completely irrelevant for a factual discussion. I remember well that people used to make fun of "giga" too before it became common. If your point is that the current articles on the issue are sub-optimal because they are too overloaded with too many criticism, discussion, repetitive information, inconsistent use of prefixes and that the same is unnecessarily duplicated in related articles, I'd agree. --NotSarenne 17:24, 3 November 2007 (UTC)

220 KiB and counting

"This page is 220 kilobytes long. It may be helpful to move older discussion into an archive subpage. See Help:Archiving a talk page for guidance."

What kind of kilobytes are those? (Answering the question is not the point. It's unfortunate that we raise the question every time we publish such a message.)

-- Pelavarre 13:28, 3 November 2007 (UTC)

It was, indeed, 220 KiB (and not 220 kilobytes) when your wrote that as one can verify by looking at the history which shows the size in bytes without use of prefixes. That's a good catch. It would be neat to either fix the message to use "KiB" or otherwise, to be "neutral" on this issue, use of prefixes should be avoided. --NotSarenne 15:23, 3 November 2007 (UTC)
It is actually kilobytes in the binary sense (225,644 bytes = 220.36 kilobytes = 220.36 KB), someone a long time ago changed the template to display KiB when the other wikipedia templates use kilobytes and KB. Fnagaton 16:16, 3 November 2007 (UTC)
Well, that's what I wrote, isn't it? That's why I also say, get rid of the unnecessary use of prefixes in this case. The article history doesn't use prefixes either and I believe, this would be appreciated regardless of how you'd like to define "kilobyte". It certainly does not help, if Wikipedia is inconsistent with itself. Actually, I dare to question the use of "byte" in this context. Text should be measured in terms of characters which is not the same in case of UTF-8. --NotSarenne 17:29, 3 November 2007 (UTC)
Not it's not because the other templates use the term "kilobytes" or "KB" in the binary sense. Therefore the correct option is to not use KiB but to use KB instead and the edit history for the template also show that is the case. Fnagaton 18:35, 3 November 2007 (UTC)

Basic English

Could someone who has studied English or is a very good English writer, explain why this revert is justified: [13]? In how far, is my edit worse than the other revision? To me "[...]decimal prefix meanings of KB[...]" really makes no sense. I've explained above what prefix is and that KB is not a prefix but a unit. --NotSarenne 20:30, 1 November 2007 (UTC)

Personally I don't think the revert is justified in that your original statement is unambiguous, while the reverted statement arguably is ambiguous in that an uninformed person might think that KB, etc., are prefixes. An informed reader would recognize that they are prefixed units and deal with the prefix separate from the unit. In the end, I doubt if the potential confusion is sufficient to justify the angst. Tom94022 03:21, 2 November 2007 (UTC)
Thanks for your response and the edit. Though I do think that this version was better: [14]. Maybe the grammar isn't quite correct? Let me explain why I consider it better: The first online mention of the term binary prefix seems to be this here: [15] by Markus Kuhn. So, strictly speaking, it refers only to KiB, MiB, GiB etc. but not to SI prefixes used in the "binary sense". The term "binary prefix" seemingly did not exist before this standard. With that in mind I'd say that your phrasing was less clear than the previous version. That's also why I think the headline of the paragraph "Binary prefixes using SI symbols" is nonsense. It should be "[something] expressed (ab)using SI symbols" or similar whereas I'm struggling with the [something]. These are not binary values or numbers. Apparently it should be "multiples of powers-of-two". --NotSarenne 02:45, 4 November 2007 (UTC)

Proposal: Keep redirects from KB, MB, GB, TB, PB, EB neutral

In my opinion it is not useful to repeat information like 1 kB may denote either 1,000 bytes or 1,024 bytes on the disambiguation pages. Otherwise, these will always additional places for disputes regarding the binary prefix notation. The reader must follow the links anyway to understand why there are different meanings/uses and which of them is used in what contexts. I think the current pages for TB is most appropriate, so KB should read like this:

* kilobit (kb), a unit of information or computer storage
* kilobyte (KB, kB), a unit of information or computer storage

Any objections? --NotSarenne 16:59, 4 November 2007 (UTC)

I support this proposal. There's already enough discussion in the individual articles - there's no point in encouraging it to spill over into the re-direction pages. Thunderbird2 17:19, 4 November 2007 (UTC)


I'd prefer to get some other opinions first. As it affects several articles, it may be wise to raise in wp:mosnum before carrying out the change. Would you object to that? Thunderbird2 18:30, 4 November 2007 (UTC)
My opinion is no longer accepted here. I just wanted to let you know, that there is no point in waiting for a response from User:NotSarenne. I'm almost certain the proposed changed would be accepted though as long as it's not me who makes them. You could certainly try to discuss it at WP:MOSNUM, if you think that's necessary. --217.87.59.247 18:35, 4 November 2007 (UTC)

Common usage and standards

There seems to be a conflict between (1) how words like megabyte and gigabyte are used by most speakers of English and (2) standards proposed or adopted by various bodies.

Is there also a US-EU split or a Windows/Linux split here?

Maybe we should indicated which language communities, regions, professions, etc. use the various gigabyte vs. gibibyte terminology.

Basically there are two meanings of the mega, giga, tera serios of prefixes: powers of 1000 (as in the metric system invented over a century ago in France) and powers of 1024 (used in recent decades mostly in reference to file size).

Nothing personal about it, I'm just reporting what I have seen and heard. --Uncle Ed (talk) 04:46, 21 December 2007 (UTC)

Interesting idea, but it sounds like Original Research. If you can find research that has already been done on this subject, and is widely published outside the Wikipedia reference frame, why not? 84.196.45.22 (talk) 11:02, 6 April 2008 (UTC)

Absurdity of binary prefixes

Hasn't everyone noticed that the so called binary prefixes are always used with decimal digits which, IMO. is quite absurd since it forces conversion factors for each change in prefix. Isn't the whole idea of prefixes is to not have conversion factors, just shift the digit point? Binary prefixes might make sense with digits in Hexadecimal or Octal but then the prefixes would not be of the form 2n*10 (n=1,2,3,etc) but two would be raised to other powers such as 212 for a 3 Hex digit shift or 29 for 3 Octal digits. Also note that a 3 digit shift doesn't make a lot of binary sense either, two or four would be much more binary. Since programmers and engineers are generally lazy, this further suggests to me that binary prefixes are never used in any serious engineering and/or programing calculations but instead are shorthands used by technical, marketing and GUI persons to save space in reports, advertisements, displays, etc. Therefore, IMO, when these prefixes were used in a binary sense, they were never intended to be precise values and should not be interpreted in such a manner. Tom94022 (talk) 00:34, 17 January 2008 (UTC)

Recently I did an example that shows exactly how much consensus the IEC "standard" has in the real world. A while ago on Wikipedia we had one user edit hundreds of articles to change from kilobyte to kibibyte (and all of the other units as well) so since this would alter any attempt to use Google to judge real world consensus on this issue the searches are conducted with "-wikipedia".
Historical use search terms Results
kilobyte -wikipedia 1,940,000
megabyte -wikipedia 6,190,000
gigabyte -wikipedia 3,640,000
Total: 11,770,000
IEC Search terms Results
kibibyte -wikipedia 28,800
mebibyte -wikipedia 17,100
gibibyte -wikipedia 19,000
Total: 64,900
Consensus for historical use: 99.449%
This shows the IEC standard does not have consensus in the real world, so I don't think the IEC can be seen as authoritative in this regard.Fnagaton 08:57, 17 January 2008 (UTC)
"the moon is made of rock" 1670
"the moon is made of cheese" 27500
"the moon is made of green cheese" 25900
Obviously, we'll need to change the article about the Moon. Though the color appears to be in dispute. Iron Condor (talk) 00:41, 17 May 2008 (UTC)


I didn't make my point clear, k = 1024, M = 1,048,576, etc are absurd! Ki, Mi, etc. happen to be an unfortunate necessity resulting from incompetent and/or lazy programmers at Apple and Microsoft using K,M, etc. in an unusual way without explaining their usage. Tom94022 (talk) —Preceding comment was added at 20:23, 17 January 2008 (UTC)

K=1024 etc is not absured. What is absured are the attempts by the IEC to rewrite what is already well known. Most of the world disagrees with the IEC. Fnagaton 21:07, 17 January 2008 (UTC)

Yr statistics are questionable since you cannot tell whether megabyte and kilobyte are used in a decimal sense or a binary sense. Pick your own term, but u never addressed my observation that the combination of decimal digits with a binary prefix doesn't make a lot of sense. Tom94022 (talk) 22:04, 20 January 2008 (UTC)

What "proportion" is irrelevant to my point, as is the use of base ten numbers with binary prefixes. My point is more than 99% of the internet uses those terms and less than 1% use the so called IEC "standard". You may think it doesn't make a lot of sense however I think it makes perfect sense. Fnagaton 23:21, 20 January 2008 (UTC)
But your statistics do not say anything to wheter those terms are used in a binary manner. Did your research exclude pages of hdd-manufacturers, access providers, network hardware manufacturers, blank DVD manufacturers and all the other guys using these prefixes in a decimal manner? I wouldn't be surprised if them amount to half of your hits. And, intentional or not, those guys support the "so called IEC standard". At least they do not support the binary meaning of SI-Prefixes. --213.183.10.41 (talk) 18:31, 11 March 2008 (UTC)
Your point is still irrelevant because the proportion of those used in the power of two sense is not important to the point I am making. The point is that the -bi terms have very limited use in the real world. Your guess of "half" is just that, a guess without any substance. Fnagaton 22:07, 17 March 2008 (UTC)

Archive time?

The length of this Talk page gives me a headache. CapnZapp (talk) 18:52, 8 February 2008 (UTC)

"The bad-ass of kibi promotion is Wikipedia."

I completely agree with this site: http://www.wandawanders.com/content/view/124/73/

There should be a prominent section on these articles making it clear this scheme isn't agreeable to all. CapnZapp (talk) 18:57, 8 February 2008 (UTC)

The website is down but the article is still in Google's cache http://www.google.com/search?q=cache:4eyBLXmTx5AJ:www.wandawanders.com/content/view/124/73/ Do you really think it's a noteworthy article? --217.87.88.179 (talk) 18:27, 10 February 2008 (UTC)
I objected strongly when the changes were made to the style manual back elsewhen, and it was clear that no actual consensus was really achieved to "force" the changes onto Wikipedia. My attitude still is that the whole issue of KiB vs. KB is nitpicking over nothing, and in fact the articles that try to push the term "kilobyte" to mean 10^3 bytes rather than 2^10 bytes is nothing more than naval gazing.
Time will tell if the proper choice was made, but it certainly doesn't meet with "current" industry practices and popular usage. I have other battles to fight over than this one, so it seems that only those with a real axe to grind and a desire to push a POV are winning here. I have patience, and ten years from now we will see if these silly terms get adopted into mainstream computer culture or not, at which time all of this nonsense can be edited out of Wikipedia for once and for all. And I'll be here ten years from now still working on Wikipedia in one form or another. --Robert Horning (talk) 06:21, 9 February 2008 (UTC)
Sorry but I don't quite understand what are you trying to say. Who is an axe grinder? In how far is this related to naval gazing exactly? What is "nonsense" and which are the "silly terms" in this context? Also what makes you certain you'll be still here in ten years and what relevance has this assumption? Please clarify. Thanks. --217.87.88.179 (talk) 18:30, 10 February 2008 (UTC)
There is nothing silly about an encyclopaedia attempting to distinguish between two different definitions of megabyte, when both definitions are in common use by the computer and communications industries. I think the term "bad-ass" is completely inappropriate, but if the blame lies anywhere, it is with the computer industry as a whole for failing to standardise on the meaning of this and related terms. Thunderbird2 (talk) 09:10, 9 February 2008 (UTC)
Wikipedia uses consensus so it is silly to try to use terms that do not have real world consensus. Since kibibyte etc do not have real world consensus and do not have consensus in Wikipedia either then it is not the place of individual editors to try to enforce use of kibibyte etc. Don't worry Robert many other editors (including myself) have caused the binary prefix entry for WP:MOSNUM to change from the version that enforced using these neologisms. It now makes it perfectly clear that "There is no consensus to use the newer IEC-recommended prefixes in Wikipedia articles to represent binary units." Fnagaton 10:43, 9 February 2008 (UTC)
At the end of this year we should look at the adoption of the IEC binary prefixes. A full decade is long enough to see if the standard will ever be adopted by the computer industry and technical press. If the adoption rate is the same as today, the manual of style should discourage the use the IEC prefixes on Wikipedia. -- SWTPC6800 (talk) 01:33, 10 February 2008 (UTC)
If Wikipedia wishes to make unambiguous statements about computer storage and data transfer rates (and in my opinion it should), then it needs to have unambiguous units in which to express such statements. Thunderbird2 (talk) 11:34, 10 February 2008 (UTC)
Which is to specify the exact number of bytes and not use the "bi-" neologisms because they are not widely used and can cause confusion. Fnagaton 12:18, 10 February 2008 (UTC)
Yes, that is a good way of removing the ambiguity. Let us hope that editors start to follow it. But the confusion is caused by the ambiguity that was there all along, not by the attempts of the IEC to resolve it. Thunderbird2 (talk) 12:51, 10 February 2008 (UTC)
If the IEC had said "kilobyte" is now defined as 1024 bytes (like the JEDEC did) that would have made more sense as it would have followed real world consensus. Much better than inventing new terms that don't used used by the vast majority of people, even after nine years. ;) Fnagaton 16:01, 10 February 2008 (UTC)
Fnagaton, there is no real world consensus on a binary meaning of SI-prefixes in conjuntion with Bits or Bytes, and it has never been. Despite JEDEC, most hardware manufacturers at least also use the decimal meaning of SI-prefixes. Some are even this schizophrenic using both meanings in one product (motherboards, cpu, graphic card: memory amount/bus rates; harddisks: magnetic memory/cache memory). You may say IEC-prefixes are uncommon or even nearly not existent. But considering a binary meaning of MB a real world consensus are visionaries. And exactly that's reflected by the result of the wikipedia consensus: there is no consensus if MB is decimal or binary. It is simply ambiguous or context sensitive at best. --213.183.10.41 (talk) 19:20, 11 March 2008 (UTC)
So if I want to talk about 2 GiB of RAM I should say 2147483648 bytes of RAM?! As for confusion, simply wikifying MiB would remove any confusion as to what a MiB is. If you prefer to use GB, the context should make it clear which one you're talking about. The gigabyte page lists when the 1024-based definition is used and when the 1000-based definition is used. Andareed (talk) 13:15, 10 February 2008 (UTC)
The suggestion is that 1 GiB would be written as 1 GB (10243 B) and the decimal gigabyte as 1 GB (10003 B). I don't know whether the context (without this clarification) would make it clear to an expert. But we are not writing for experts, and it certainly does not make it clear to me. Thunderbird2 (talk) 13:30, 10 February 2008 (UTC)
This is ridiculous imho. There's no need for the additional redundancy. People familiar with GB/GiB know what the contexts where GB = 10^9 and where GB = 2^30 (and if not they can read the articles on GB or GiB). As for experts, I doubt most people are familiar with the Kelvin temperature scale. But if you look at the article on the Sun, temperatures are written exclusively in degrees Kelvin with no Fahrenheit or Celsius equivalents given. Andareed (talk) 20:18, 10 February 2008 (UTC)
I'm equally comfortable with the use of GiB for disambiguation (as an alternative to 10243 B), but others disagree. At the moment WP:MOSNUM#Binary_prefixes permits both styles. Thunderbird2 (talk) 13:38, 10 February 2008 (UTC)
Andareed, using -bi (especially with it wikified) as the main unit adds confusion because in the majority of cases it changes the article units compared to the units found in the sources relevant to the article. Consistency with real world language and hence consensus is important. I also think using -bi as disambiguation adds confusion by introducing less well known terms than is expected by common use and that found in real world consensus. Your solution, wikifying it, can be equally applied to the existing terms kilobyte/megabyte/etc to make it clear to the reader how many bytes it is and this way it doesn't introduce an extra neologisms. For example if the sections existed in the pages then using something like [[kilobyte#Binary use|kilobyte]] or [[kilobyte#Decimal use|kilobyte]] would work. Fnagaton 15:59, 10 February 2008 (UTC)
What is the meaning of "confusion"? Exactly nobody is confused when they read MiB or Mebibyte. Not a single person. Some people confuse "confusion" with "lacking knowledge of terminology". What does neologism mean? Are neologisms something bad? It's derived from ancient greek, neo means new and logos means word. kilo is also ancient Greek and the word kilo means thousand. It's not a new word. Shouldn't the English Wikipedia use English instead? Is anyone here from ancient Greece? Mebibyte is a relatively new word but Megabyte isn't exactly old either but every educated person knows that "mega" means million and its definition hasn't changed for centuries. --217.87.88.179 (talk) 22:25, 10 February 2008 (UTC)
Confusion - The "lack of clearness or distinctness" or "the state of being confused". It easily demonstrated that you are wrong what you write "Exactly nobody is confused when they read MiB or Mebibyte" with this web page where you will note it is claimed that "1 KiB = 1000 bytes and 1 MiB = 1000000 bytes". Neologism is "a new word, meaning, usage, or phrase" and is perfectly valid English. If you look at the JEDEC standard then you have "kilo (K) (as a prefix to units of semiconductor storage capacity): A multiplier equal to 1024 (210)." and "mega (M) (as a prefix to units of semiconductor storage capacity): A multiplier equal to 1,048,576 (220 or K2, where K = 1024)." which refutes your claim. Fnagaton 22:35, 10 February 2008 (UTC)
So if someone accidently writes 1*1 = 2, that proves there are people who believe that 1*1 is literally 2? Wouldn't someone agree that it's far more likely this person just made typing mistake? In any case it's exactly as I wrote. This person isn't confused. This person did not know what KiB meant and had to look it up. A pity isn't it? A person who does not know everything. How can this be? JEDEC has exactly zero authority and they can speak for a very tiny fraction of the industry only anyway. The IEEE and other many other international standards organisations have authority, they have adopted the IEC standard and encourage its use. Kibibyte is also a valid English word. --217.87.88.179 (talk) 22:51, 10 February 2008 (UTC)
Your 1*1=2 example is irrelevant because it does not tackle the actual issue and because the example I gave goes way beyond a simple typo. You are wrong because JEDEC is authoritative and "is the leading developer of standards for the solid-state industry". The supporters of the -bi prefixes can be said to not have authority by the simple fact that the terms -bi have only ~0.5% use in the real world after nine years. Also ANSI/IEEE Std 1084-1986 says "kilo (K). ... In statements involving size of computer storage, a prefix indicating 210, or 1024" and "mega (M). ... In statements involving size of computer storage, a prefix indicating 220, or 1,048,576". Lastly, your statement about "kibibyte is valid English" is ignoratio elenchi. Fnagaton 23:08, 10 February 2008 (UTC)
My example is just as irrelevant as yours. JEDEC maybe the "self-proclaimed" whatever but this is like-wise irrelevant. Wikipedia is not a semi-conductor. So what JEDEC believes is of no concern. Again, you misinterpret Google results. This has been explained to you more than once and your 0.5% figure is absolutely irrelevant. You also forgot to mention that you're quoting a deprecated standard. The IEEE has revised this standard and adopted the IEC binary prefixes. They are the status quo whether people use them or not. --217.87.88.179 (talk) 23:24, 10 February 2008 (UTC)
No your example is not relevant for the same reasons given above, your example also relies on a false premise. I also note you don't offer any evidence to support your claims, your claims are incorrect so it is no wonder you cannot support them with reliable evidence. The "standard" you mention is at best a failed standard since it is not widely accepted, it is obviously not the "status quo" since the state of affairs in the vast majority of the industry is to not use those prefixes. I am not misinterpreting the Google results (I also note you do not show how, instead you just make a baseless claim) the 0.5% real world consensus is directly relevant to this subject since Wikipedia:Verifiability is official policy here. Fnagaton 23:36, 10 February 2008 (UTC)
You want me to explain why Google results cannot be used to determine how wide-spread certain words are? Are you serious? Standards cannot fail because standards don't compete. Products based on standards may fail because the products compete on the market but standards exist or do not exist. It's no race. Standards can be revised or deprecated. The IEC binary prefixes have not been deprecated. Your example relies on a false premise. The author of the public mail writes "The MiB (and its derivatives) was invented by hard drive manufacturers in a very lame attempt to make their drives capacity look bigger". This could not be further from the truth and everybody here knows that very well. This person is not confused at all. He's trying to create confusion by intentionally misusing these units. The mail is from "January 24, 2008". A bit late for confusion especially if the very same mail contains a software patch to make use of these units. --217.87.88.179 (talk) 23:59, 10 February 2008 (UTC)
(out dent) The Google test is a very good and accepted way to show how often certain words are used in the real world, especially words like kilobyte/kibibyte because those words are not likely to be used by other topics except the subject they are related to. Also this "anonymous" IP user from your ISP who writes on this same subject used the Google test. It is funny how you claim the Google test "cannot be used" when is refutes your argument. The JEDEC standard, which is authoritative, trumps the other so called standards you support. Now you are making conclusions "He's trying to create confusion by intentionally misusing these units" where you don't supply any supporting evidence, so it is illogical for you to state what you just did. Like I said, your argument relies on a false premise, actually now it is more accurate to say your argument relies on more than one false premise. Fnagaton 00:15, 11 February 2008 (UTC)
The current Google test is 1-dimensional. It should be 2-dimensional, with the second dimension being time. I mean: if you do a simple Google search, you get all search result accumulated over time, in this case over the last 10 years. A more correct Google test would be to divide the results in 10 classes, one for each year, to see the evolution over time of the usage or non-usage of certain words.

I am not a Google expert so I don't know how to do this. 84.196.45.22 (talk) 11:24, 6 April 2008 (UTC)

It is not accepted at all. I'm almost certain it's even mentioned in some Wikipedia policy. The internet and especially the www are only a tiny fraction of "the real world" and they are not representative. For most people, there cannot be found any evidence of their existence anywhere on the web at all and even Google indexes only a fraction of the web, not to mention that many information is copied and repeated like mailing list archives which leads to very wrong numbers, very quickly but the ratio differs for each and every single case, except that older information is more likely duplicated than newer. Regarding "intentional misuse", it is obvious that there are only two possibilities: intentional or non-intentional misuse. You claim that the person is confused. I claim he isn't because you don't go ahead and write a software patch to use KiB/MiB/GiB without knowing what these units mean. The patch looks actually correct, so he does not seem to be confused after all. So if he didn't mistype 1000 instead of 1024 in his mail, he must have done it on purpose. Or do you think he suffers from multiple personality disorder? This would be hard to prove. After all it's much more likely that he thought one thing and typed another. These kind of things happen. Do I have to prove it with Google? So we're back to zero. There is still not a single person confused about the meaning of "KiB". It's just that some people don't know what it means and some may pick the wrong unit on accident. --217.87.88.179 (talk) 00:50, 11 February 2008 (UTC)
Of course it is accepted, but it just happens to show you are wrong. Your point is fallacious because there is no reason to suggest that Google gives more weight to indexing kilobyte over kibibyte, so the terms from the search results therfore show the proportion of the consensus in the real world. Your statement about "you don't go ahead" and the follwoing statements about the software patch contradicts itself. Your later claim about "multiple personality disorder" is irrelevant and shows that you have not provided any valid supporting argument. The bit about "don't know what it means and some may pick the wrong unit" is actually one definition of being confused, so you refute yourself. So to sum up you are wrong because some poeple are confused about the units as demonstrated by the link I supplied. Fnagaton 01:02, 11 February 2008 (UTC)

From Wikipedia:Verifiability The threshold for inclusion in Wikipedia is verifiability, not truth.

It doesn't matter if kibi is more accurate. The "Reliable sources" in the computer industry and technical press do not use the IEC binary prefixes. A few standards organizations have proposed a new method of measuring binary storage capacities but the industry said "No thanks". -- SWTPC6800 (talk) 22:11, 10 February 2008 (UTC)


I don't understand why you guys are wasting your time with a previously multiply blocked and banned user evading under multiple ip's - User:Sarenne, User_talk:NotSarenne, User_talk:217.87.59.247, etc. etc. etc. and now the same dialup location, Special:Contributions/217.87.88.179 --Marty Goldberg (talk) 23:13, 10 February 2008 (UTC)


I'm thinking maybe an sprotect is in order for this page for the time being? --Marty Goldberg (talk) 17:35, 11 February 2008 (UTC)

Yes definitely. Users Sarenne/NotSarenne certainly love to hop IPs in their ISP. Fnagaton 17:37, 11 February 2008 (UTC)
 Y Semi-protected temporarily. — Satori Son 18:44, 11 February 2008 (UTC)
After reading all of these replies, I have only one argument. Wikipedia should not spearhead, it should follow. It should reflect general usage, not try to change it. Let's hope we are allowed to change back all those kibibytes soon, and get rid of the confusion and embarrassment ("why is Wikipedia so elitist and weird?") once and for all. CapnZapp (talk) 22:15, 1 June 2008 (UTC)