Wikipedia talk:Manual of Style (dates and numbers)/Archive B9

This page is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Proposed change to wording of "Binary prefix" (2/3): search for consensus

Attempt to find out where everyone stands on separate points

So, here's some statements. Under each one, it would be useful if anyone who feels themselves involved in this debate indicate whether they agree or disagree with the statement. Hopefully, they';ll be no need in this section to say more than that; hopefully we can leave subtle distinctions for now, and caveats will be covered by the other points. However, if you really need to say more, then that could make sense; there's just no need to explain why you agree or disagree here. I've indicated my own views through these. If you want to add an extra statement, go right ahead. Just to note, this time I'm not asking people to judge where there is or isn't consensus, just to give their own views.

The rationale here is that I'm fairly convinced that we've gotten tangled up and don't realise which things we already all (or nearly all) agree upon. If you don't want to pick "agree" or "disagree", feel free to say "undecided". SamBC(talk) 18:17, 27 March 2008 (UTC)

These statements are not sufficiently unambiguous and are inviting applying improper conclusions. --217.87.90.29 (talk) 19:49, 27 March 2008 (UTC)

Really? Part of the idea was that they aren't (all) mutually exclusive, so subtleties can be indicated by combinations; but I feel that, if they're read for no more than what they say, they're pretty clear. Could you give a counterexample?

Some of the comments added confirm that these statements are problematic. For example, the sentence "IEC units should not be used in the general text of an article, but are an ideal choice for disambiguation" combines two statements. This is certainly valid but complicates things and people may agree on this statement but still disagree on the meaning because they understand them differently or apply differing emphasis. I don't see the necessity to combine these two and the use of "ideal" is not ideal either. The statements added by Thunderbird2 are better (albeit not perfect either) because they are shorter and straight-forward. I find it perfectly valid to point out flaws without providing better alternatives, as long as the arguments are valid. --217.87.66.130 (talk) 16:22, 29 March 2008 (UTC)

IEC units shall not routinely be used, except in quotes, articles specifically discussing the units, and articles in which they are employed by the primary cited source

Question in the interests of comparability, can we read this as being equivalent to the "should be banned" wording, or the "should be discouraged" wording, or neither; if neither, how do you mean it to differ? SamBC(talk) 20:07, 27 March 2008 (UTC) The new title is as close to what the proposal actually calls for as I can make it. As this is essentially voting on the proposal, I think it is fair to assume that the votes below, if everyone weighed in here, would reflect what is seen in the above voting. Greg L (my talk) 23:12, 28 March 2008 (UTC)
Agree Greg L (my talk) 18:28, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 18:40, 27 March 2008 (UTC)
Disagree Jeh (talk) 20:48, 27 March 2008 (UTC)
Disagree Tom94022 (talk) 05:11, 28 March 2008 (UTC)
Agree Fnag aton 16:41, 28 March 2008 (UTC)
Disagree LeadSongDog (talk) 23:47, 28 March 2008 (UTC)
Disagree Dpbsmith (talk) 23:50, 28 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)

By the way, if you click on this link: [edit], you can vote in all these segments in a single session. Greg L (my talk) 22:51, 28 March 2008 (UTC)

Disagree, mathematical conversions to different units of measurement are always allowed, and well that they should be, here and elsewhere. The use of an ambiguous unit over an unambiguous one does readers a tremendous disservice. Seraphimblade ^{Talk to me} 15:07, 29 March 2008 (UTC)
Disagree, obviously. — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree, for reasons in sections below. — Omegatron 05:47, 6 April 2008 (UTC)

IEC units should be banned from wikipedia outright, except where policy would require their use

Disagree SamBC(talk) 18:17, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 18:40, 27 March 2008 (UTC)
Agree (except where policy would require {or permit} their use, which is similar to the above) Greg L (my talk) 18:28, 27 March 2008 (UTC)
Disagree Jeh (talk) 20:48, 27 March 2008 (UTC)
Disagree Tom94022 (talk) 05:11, 28 March 2008 (UTC)
Agree Fnag aton 16:41, 28 March 2008 (UTC)
Disagree LeadSongDog (talk) 23:47, 28 March 2008 (UTC)
Disagree Dpbsmith (talk) 23:59, 28 March 2008 (UTC)
Disagree — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree — Omegatron 05:48, 6 April 2008 (UTC)

IEC units should be banned except for direct, verbatim quotes, and articles discussing the units themselves

Disagree SamBC(talk) 18:17, 27 March 2008 (UTC)
Disagree (partial solution) Greg L (my talk) 18:32, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 18:41, 27 March 2008 (UTC)
Disagree Jeh (talk) 20:48, 27 March 2008 (UTC)
Disagree Tom94022 (talk) 22:31, 28 March 2008 (UTC)
Disagree LeadSongDog (talk) 23:49, 28 March 2008 (UTC)
Disagree Dpbsmith (talk) 23:59, 28 March 2008 (UTC)
Disagree, proper use should be encouraged. Seraphimblade ^{Talk to me} 15:05, 29 March 2008 (UTC)
Disagree — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree DavidPaulHamilton (talk) 23:31, 1 April 2008 (UTC)
Disagree — Omegatron 05:49, 6 April 2008 (UTC)

IEC units should be discouraged, except where they are widely used by the sources for an article

Question What I mean is that there's a world of difference between deprecating IEC prefixes and encouraging alternatives. Which is intended here? Thunderbird2 (talk) 23:27, 27 March 2008 (UTC)
- Answer My dictionary defines "discourage" as "show disapproval of". Thunderbird2 (talk) 23:40, 27 March 2008 (UTC)
- Answer the general meaning of this is meant to be basically the same as the extra point added by Greg. It's probably my worst-worded point, actually. Basically, by discourage I was meaning to indicate that we might say "don't use these without some good reason". Like "should not generally be used", followed by the exceptions that are seemingly accepted by even the most anti-IEC participants here (ie direct quotes, articles talking about the IEC units, and articles whose sources use IEC). SamBC(talk) 00:14, 28 March 2008 (UTC)
Agree SamBC(talk) 18:17, 27 March 2008 (UTC)
Conditionally If the proposal isn’t adopted, this is a good first step. Greg L (my talk) 23:15, 28 March 2008 (UTC)
Undecided My position on this depends on how you define "discouraged". Thunderbird2 (talk) 18:46, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 23:46, 27 March 2008 (UTC)
Disagree Jeh (talk) 20:48, 27 March 2008 (UTC)
Disagree Tom94022 (talk) 05:12, 28 March 2008 (UTC)
Agree Fnag aton 16:42, 28 March 2008 (UTC)
Disagree LeadSongDog (talk) 23:51, 28 March 2008 (UTC)
Disagree Dpbsmith (talk) 00:00, 29 March 2008 (UTC)
Disagree, conversion of units of measurement is a simple, factual, indisputable mathematical task and has never been forbidden, even if sources may use a different measurement type. There's no difference here. Seraphimblade ^{Talk to me} 15:04, 29 March 2008 (UTC)
Disagree — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree — Omegatron 05:50, 6 April 2008 (UTC)

IEC units are good, and should be used wherever feasible

Comment Would switch to agree if "should be used ..." were replaced with "may be used where appropriate". Thunderbird2 (talk) 10:07, 29 March 2008 (UTC)
Disagree SamBC(talk) 18:17, 27 March 2008 (UTC)
Disagree Greg L (my talk) 18:32, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 18:48, 27 March 2008 (UTC)
Agree Jeh (talk) 20:48, 27 March 2008 (UTC)
Disagree I would agree if the work feasible were changed to appropriate.Tom94022 (talk) 05:14, 28 March 2008 (UTC)
Agree if word "feasible" is changed to "appropriate", as with Tom94022. Dpbsmith (talk) 23:50, 28 March 2008 (UTC)
Agree if word "feasible" is changed to "appropriate", as with Tom94022. LeadSongDog (talk) 23:53, 28 March 2008 (UTC)
Agree for the symbols at least. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree personally, but recognize that some other Wikipedians abhor them. "Appropriate" is better for compromise purposes. — Omegatron 05:53, 6 April 2008 (UTC)

IEC units should not be used in the general text of an article, but are an ideal choice for disambiguation

Disagree 18:17, 27 March 2008 (UTC)
Disagree Greg L (my talk) 18:32, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 18:49, 27 March 2008 (UTC)
Disagree IEC units SHOULD be used in the general text, that is the whole point. Jeh (talk) 20:48, 27 March 2008 (UTC)
Disagree Tom94022 (talk) 05:14, 28 March 2008 (UTC)
Disagree Fnag aton 16:42, 28 March 2008 (UTC)
Disagree LeadSongDog (talk) 23:55, 28 March 2008 (UTC)
Disagree Dpbsmith (talk) 00:00, 29 March 2008 (UTC)
Disagree, why not use a well-defined term when appropriate? Seraphimblade ^{Talk to me} 15:02, 29 March 2008 (UTC)
Undecided, because I disagree with the first part and agree to the second. — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree - Should be used in general text of the article. Relegating disambiguation to footnotes is a poor solution. — Omegatron 05:54, 6 April 2008 (UTC)

SI-style byte multiples should be disambiguated as appropriate

Agree SamBC(talk) 18:17, 27 March 2008 (UTC)
Agree Thunderbird2 (talk) 18:38, 27 March 2008 (UTC)
Disagree Greg L (my talk) 18:32, 27 March 2008 (UTC)
Disagree if an SI style multile needs disambiguating that only proves that the IEC version should have been used. Jeh (talk) 20:48, 27 March 2008 (UTC)
Agree Most people have no idea what MB means, the next largest category thinks it means million bytes, only programmers and the like know it might mean something about 1,048,000 (I refuse to calculate the number any more)Tom94022 (talk) 05:17, 28 March 2008 (UTC)
Agree But not disambiguated with IEC prefixes since that doesn't solve the problem, it adds to the confusion by adding virtually unknown terms. Fnag aton 16:44, 28 March 2008 (UTC)
Disagree If the appropriate units are used, no dab is needed LeadSongDog (talk) 23:58, 28 March 2008 (UTC)
Agree. What matters is whether the reader understands what it meant. Dpbsmith (talk) 00:01, 29 March 2008 (UTC)
Agree, if “as appropriate” means if they are used in a binary sense or when there is profound reason to think they could be. OTOH, Jeh has a point, too. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree --agr (talk) 20:02, 2 April 2008 (UTC)
Disagree personally - Ideally, SI prefixes would be used consistently throughout the encyclopedia, with no exceptions made for computing, and this wouldn't be necessary. Obviously many people want to use the units they are familiar with, however. — Omegatron 06:00, 6 April 2008 (UTC)

SI-style byte multiples should be disambiguated at every occurrence, except close repetitions

Disagree SamBC(talk) 18:17, 27 March 2008 (UTC)
Disagree Greg L (my talk) 18:32, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 18:37, 27 March 2008 (UTC)
Disagree per above comment. Jeh (talk) 20:48, 27 March 2008 (UTC)
Agree per above comment, disambiguation is necessary and this could easily be done by defining it as a link each time, e.g. MB and cleaning up the destination pages. Tom94022 (talk) 05:22, 28 March 2008 (UTC)
Agree Dpbsmith (talk) 00:01, 29 March 2008 (UTC)
Disagree if dab means more than a wikilink to the appropriate unit. LeadSongDog (talk) 00:20, 29 March 2008 (UTC)
Disagree. Cluttering is to avoid for readability reasons. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree Once per article is enough in most cases. This is at most a technical nit of little interest to most readers who aren't already aware of the distinctions.--agr (talk) 20:04, 2 April 2008 (UTC)
Kind of agree - links aren't really that cluttery, but every occurrence should not be linked; just ones that matter, like the first in the article, the first in a section discussing some different topic, etc. — Omegatron 06:03, 6 April 2008 (UTC)

SI-style byte multiples do not need disambiguation where it is clear that one is speaking about RAM, HDD, etc, where the meaning is generally "obvious"

Undecided SamBC(talk) 18:17, 27 March 2008 (UTC)
Disagree Greg L (my talk) 18:32, 27 March 2008 (UTC)
Disagree Thunderbird2 (talk) 18:36, 27 March 2008 (UTC)
Disagree It is "obvious" to me that an 80 GB hard drive is 80,000,000,000 bytes, but not to many consumers, and most operating systems further the confusion by reporting it as "75 GB" or so. Jeh (talk) 20:48, 27 March 2008 (UTC)
Disagree Most consumers don't have a clue. They become confused when the operating system reports a different number for a HDD than is labeled on the drive and its box. I also bet they also get confused when they add a 512 MB RAM module and they find the capacity increase is not 512,000 KB. Tom94022 (talk) 05:28, 28 March 2008 (UTC)
Agree Fnag aton 16:44, 28 March 2008 (UTC)
Strongly disagree. I would have agreed a few decades ago when the usage really was consistent. That's not true any more. Usage has changed over time and is, at the present time, inconsistent. Lawsuits over drive capacities show that this is not a moot point. You cannot infer the meaning from context. Dpbsmith (talk) 00:03, 29 March 2008 (UTC)
Disagree LeadSongDog (talk) 00:23, 29 March 2008 (UTC)
Agree. MOSNUM should say where which one is the default. (Actually I’d of course still prefer them to be decimal all the time.) — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree as long as we don't have a rule mandating standard SI prefixes everywhere. There really isn't a consistent "common usage" otherwise, and if there were, it would not be "obvious" to our readers anyway. Just because I know the difference in measurements between CDs and DVDs doesn't mean the general population does. They should be disambiguated in some way at least once. — Omegatron 06:09, 6 April 2008 (UTC)

I added a few more theses with the same objective and following similar principles to SamBC. For "megabyte", read "kilobyte, megabyte etc" and similarly for "mebibyte" Thunderbird2 (talk) 17:18, 28 March 2008 (UTC)

The word “megabyte” (symbol MB) is ambiguous

Agree Thunderbird2 (talk) 16:40, 28 March 2008 (UTC)
Disagree Fnag aton 16:46, 28 March 2008 (UTC)
Agree --Gerry Ashton (talk) 17:31, 28 March 2008 (UTC)
Agree Tom94022 (talk) 22:33, 28 March 2008 (UTC)
Agree Greg L (my talk) 23:02, 28 March 2008 (UTC)
Agree Dpbsmith (talk) 00:04, 29 March 2008 (UTC)
Agree LeadSongDog (talk) 00:25, 29 March 2008 (UTC)
Agree SamBC(talk) 12:48, 29 March 2008 (UTC)
Agree for common language. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree Jeh (talk) 02:54, 31 March 2008 (UTC)
Fact - a "100 MB" hard drive is 100,000,000 bytes, "128 MB" of memory is 134,217,728 bytes, and a "1.44 MB" floppy is 1,474,560 bytes. — Omegatron 06:13, 6 April 2008 (UTC)
- Comment actually that "100 MB" hard drive is a bit more than 100,000,000 bytes. ;) Jeh (talk) 08:52, 6 April 2008 (UTC)

The word “mebibyte” (symbol MiB) is ambiguous

Agree As shown some people have been confused, therfore it can be ambiguous. It is also virtually unused which only adds to the potential for readers finding it to be ambiguous. Fnag aton 16:47, 28 March 2008 (UTC)
Disagree Thunderbird2 (talk) 16:40, 28 March 2008 (UTC)
Disagree --Gerry Ashton (talk) 17:31, 28 March 2008 (UTC)
Disagree Just because some folks get confused doesn't mean the term is ambiguous. Tom94022 (talk) 22:34, 28 March 2008 (UTC)
Disagree It may not be “ambiguous” as to its definition, but it is poorly understood and therefore is a very poor tool for communicating to the intended audience. Greg L (my talk) 23:02, 28 March 2008 (UTC)
Disagree It's not ambiguous. It is unfamiliar. A case can be made for defining it on its first appearance in an article, even though you'd think that Wikilinking the first appearance of the term would be good enough. Dpbsmith (talk) 00:05, 29 March 2008 (UTC)
Disagree per dpbsmith LeadSongDog (talk) 00:26, 29 March 2008 (UTC)
Disagree unfamiliar, and occaisionally misused, but well-defined with no variation due to context or whim, so not ambiguous. SamBC(talk) 12:49, 29 March 2008 (UTC)
Disagree, I doubt that your average person knows the precise definition of a mole, either, but it does have a precise definition and thus is unambiguous. The same is true of the binary prefixes. The very purpose of the binary prefixes is to remove the ambiguity inherent in having "kilobyte", "megabyte", etc., mean two different things, one of those things being different than the standard meanings that the average person expects of the metric decimal prefixes. Seraphimblade ^{Talk to me} 14:42, 29 March 2008 (UTC)
Disagree — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree Jeh (talk) 02:55, 31 March 2008 (UTC)
Disagree - Confusion is not the same as ambiguity. — Omegatron 06:14, 6 April 2008 (UTC)

The word “mebibyte” (symbol MiB) is not widely recognized by the typical Wikipedia reader

Agree Greg L (my talk) 23:02, 28 March 2008 (UTC)
Agree, but it is readily understood from context. Or Wikilinking. Or defining them on their first apperance. The people who express a strong visceral reaction to the terms are not confused by them, they merely dislike them. Dpbsmith (talk) 00:07, 29 March 2008 (UTC)
Agree per dpbsmith LeadSongDog (talk) 00:27, 29 March 2008 (UTC)
Agree Fnag aton 06:00, 29 March 2008 (UTC)
Agree Thunderbird2 (talk) 09:56, 29 March 2008 (UTC)
Agree – dpbsmith's points make sense also... SamBC(talk) 12:50, 29 March 2008 (UTC)
Partially agree, the typical reader or at least the typical computer savvy reader should be able to guess it’s something similar to “megabyte” (MB), especially for the symbol. — Christoph Päper 16:41, 29 March 2008 (UTC)
- Comment So… As long as a reader can “guess” what something means, it is appropriate for an encyclopedia to use units of measure a well-read reader has never encountered before (and likely won’t again encounter after they leave Wikipedia)? We should be able to find a better MOSNUM policy that isn’t founded on that principle. Greg L (my talk) 03:34, 31 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree TONY (talk) 07:41, 30 March 2008 (UTC)
Agree per dpbsmith and Crissov, also Seraphimblade's comment re 'mole' above. To put it bluntly, it may be unfamiliar but that is not an overriding concern here. Jeh (talk) 03:00, 31 March 2008 (UTC)
Agree I think that is an overriding concern here. The geeks in our audience can give the lecture from memory; the non-geeks don't care, don't need to know and are only confused by notation that differs for the documentation the normally use.--20:08, 2 April 2008 (UTC)
Agree, though this isn't a problem. — Omegatron 06:16, 6 April 2008 (UTC)

The word “megabyte” (symbol MB) is widely recognized by the typical Wikipedia reader

Agree Fnag aton 06:29, 29 March 2008 (UTC)
Agree Thunderbird2 (talk) 09:57, 29 March 2008 (UTC)
Depends on intent, if you mean "has heard the term before", I certainly agree. However, I would venture a guess that the average Wikipedia reader does not know that the term is ambiguous and may mean different things in different contexts. Generally, only those with a strong computing background would know this, and that is not the majority of Wikipedia's readership (nor are those the people who would benefit from disambiguation). Seraphimblade ^{Talk to me} 14:45, 29 March 2008 (UTC)
Agree on recognized, wouldn’t agree on understood, i.e. I agree with Seraphimblade. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree, caveat as per Crissov and Seraphimblade. SamBC(talk) 19:50, 29 March 2008 (UTC)
Agree, but possibly misunderstood. Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree per Crissov and Seraphimblade. "Widely recognized" does not mean "correctly understood" Jeh (talk) 03:01, 31 March 2008 (UTC)
Agree Greg L (my talk) 04:01, 31 March 2008 (UTC)
Agree - it's familiar, but that doesn't mean they know the definition. — Omegatron 06:17, 6 April 2008 (UTC)

The word “megabyte” (symbol MB) is not widely understood by the typical Wikipedia reader

Comment Based upon my brief and very unscientific survey, I am of the opinion that the typical reader does not even know that mega could be 1,000,000 much less something else. Mega means large to all but technical persons, arguably a small portion of the Wikipedia audience and therefore not typical. Accordingly, any arguments and corresponding remedies based upon lack of knowledge of a prefix or symbol should equally apply to MB and MiB, etc. Tom94022 (talk) 06:21, 29 March 2008 (UTC)

Agree Tom94022 (talk) 06:21, 29 March 2008 (UTC)
Disagree This knowledge even passes the "my Dad test". Fnag aton 06:28, 29 March 2008 (UTC)
Mu this is a subtle and complex question that depends on your precise definition of 'understood'. SamBC(talk) 13:00, 29 March 2008 (UTC)
Comment It also depends on the definition of "typical Wikipedia reader". Thunderbird2 (talk) 14:38, 29 March 2008 (UTC)
Agree, any reasonably educated person with some knowledge of the metric system will (quite reasonably) assume that the prefix "mega" means one million, as it does in megaton and similar terms. Only those with strong backgrounds in computing will know that in computer science it is sometimes used to represent a value other than one million. Seraphimblade ^{Talk to me} 14:47, 29 March 2008 (UTC)
Agree — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree—Yup, the Mom test even works. TONY (talk) 07:41, 30 March 2008 (UTC)
Question I seriously don't understand what people are "agreeing" or "disagreeing" on here. SamBC makes an important point that has not been addressed. Does "understand megabyte" mean:
- "know that a megabyte is a unit used to measure the size of large files on my computer"?
- "know that a megabyte is roughly a million bytes"?
- "know that a megabyte is exactly a million bytes"?
- "know that a megabyte is exactly 1024² bytes"?
- "know that a megabyte is sometimes a million bytes, sometimes 1024² bytes depending on the context"?
We could add 5 more theses to clarify this, but perhaps the originator of the statement (I'm not sure who it was) can clarify the intended meaning. Thunderbird2 (talk) 09:38, 30 March 2008 (UTC)
Agreee in that it is not usually correctly understood as long as it is commonly used in an ambiguous way Jeh (talk) 03:03, 31 March 2008 (UTC)
Disagree If a unit of measure that is universally used in virtually all advertisements, brochures, owners manuals, and computer magazines isn’t understood by the typical reader of a Wikipedia article on computer technology (after the usual disambiguation is applied when it applies to hard drives), then there’s no point for Wikipedia to even exist since the logical implication would be that its typical reader is mentally retarded. That is clearly not the case so I question the need for continuing a practice that amounts to we Wikipedia authors trying to save the computer industry and all its readers from themselves by “showing them the way.” Greg L (my talk) 03:55, 31 March 2008 (UTC)
Agree - I've surveyed people on this. Most Americans haven't a clue. Most Europeans assume it's used consistently like any other metric prefix. It's only the technically-oriented that know that it has multiple definitions. — Omegatron 06:20, 6 April 2008 (UTC)

Disambiguation of “megabyte” should be encouraged

Agree Thunderbird2 (talk) 17:24, 28 March 2008 (UTC)
Agree Tom94022 (talk) 22:36, 28 March 2008 (UTC)
Agree Greg L (my talk) 23:02, 28 March 2008 (UTC)
Agree Dpbsmith (talk) 00:07, 29 March 2008 (UTC)
Agree to the extent of a wikilink LeadSongDog (talk) 00:28, 29 March 2008 (UTC)
Undecided, this partly depends on whether it is being used inconsistently SamBC(talk) 13:02, 29 March 2008 (UTC)
Agree, it should always be clear whether "megabyte" is referring to the value 10⁶ or 2²⁰. This is well done by using "megabyte" or "MB" to refer only to 10⁶, and using "mebibyte" or "MiB" to refer to 2²⁰, eliminating ambiguity in terminology. Wikilinks are appropriate to serve readers unfamiliar with the difference. Seraphimblade ^{Talk to me} 14:49, 29 March 2008 (UTC)
Disagree Fnag aton 16:06, 29 March 2008 (UTC)
Agree, where necessary binary or decimal should be prepended. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree' once, in an unobtrusive way such as a wikilink--agr (talk) 20:15, 2 April 2008 (UTC)
Agree if we can't decide to use SI consistently everywhere. — Omegatron 06:25, 6 April 2008 (UTC)

Disambiguation of “mebibyte” should be encouraged

Agree Since it is virtually unknown then most readers will find it confusing, so disambiguation is needed. Fnag aton 17:37, 28 March 2008 (UTC)
Disagree Thunderbird2 (talk) 17:24, 28 March 2008 (UTC)
Disagree However, because it is relatively unknown, its use should always include a hot link to an explanation. That is, it is sufficient to use it as MiB. Same thing for KiB, GiB. We hot link everything we can, it is the perfect solution for these terms since they are unambiguous! Tom94022 (talk) 22:40, 28 March 2008 (UTC)
Disagree as per Tom94022. Jeh (talk) 03:37, 6 April 2008 (UTC)
~~Disagree~~ Agree Even though the term should only be encountered in articles where it is widely recognized by the intended audience (or in articles that describe the IEC units directly), the unit should be hyperlinked to the article describing it and should have a first-occurrence footnoting (or similar method of disambiguation); particularly so given that it is very poorly recognized by the larger computer article readership. Greg L (my talk) 04:38, 31 March 2008 (UTC)
Disagree It's not ambiguous, so how can you disambiguate it? Defining it on first appearance might be reasonable. Dpbsmith (talk) 00:08, 29 March 2008 (UTC)
Disagree just link it LeadSongDog (talk) 07:40, 30 March 2008 (UTC)
Disagree – define or link, can't disambiguate something that isn't ambiguous SamBC(talk) 13:03, 29 March 2008 (UTC)
Disagree, definition is desirable and can be established through appropriate wikilinking, but disambiguation is not desirable for a term which has no ambiguity in the first place. Seraphimblade ^{Talk to me} 14:50, 29 March 2008 (UTC)
Disagree, explanation by linking to binary prefix or similar, though. — Christoph Päper 16:41, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree - it's inherently unambiguous. The first instance should be linked just like any other topic. — Omegatron 06:26, 6 April 2008 (UTC)

The best method to disambiguate “megabyte” is to specify an exact number of bytes

Agree Or by using power notation for brevity, for example 2¹⁰Fnag aton 16:49, 28 March 2008 (UTC)
Disagree Thunderbird2 (talk) 22:11, 28 March 2008 (UTC)
Disagree This is an acceptable method. Tom94022 (talk) 22:43, 28 March 2008 (UTC)
Disagree Different methods of disambiguation are needed in different situations; there is no single “right” way. Greg L (my talk) 23:02, 28 March 2008 (UTC)
Disagree it isn't best, just clearest. LeadSongDog (talk) 00:34, 29 March 2008 (UTC)
Disagree no best method, lots of factors to weigh up SamBC(talk) 13:04, 29 March 2008 (UTC)
Disagree, while it's better than no disambiguation at all, most readers would find this confusing and be unable to relate it to anything. The binary terms and abbreviations are close to the decimal terms, so a reader would (correctly) intuit that "MiB" or "mebibyte" is at least near the value of a megabyte, while likely being confused upon seeing a purely numerical representation. (If used as a secondary method, however, such disambiguation could possibly be useful in some cases.) Seraphimblade ^{Talk to me} 14:52, 29 March 2008 (UTC)
Disagree — Christoph Päper 16:41, 29 March 2008 (UTC)
Neutral Woodstone (talk) 07:11, 30 March 2008 (UTC)
Neutral This is acceptable, but I think there are better ways.--agr (talk) 20:17, April 2008 (UTC)
Disagree in general. There may be occasions where this is the best method but they will be uncommon. For a hard drive size (one of the cases where the binary/decimal prefix issue is the most problematical) do you use the exact meaning of the SI prefix (e.g. "750,000,000,000 bytes" or do you use the exact hard drive size? Which will only be valid for one particular make and model of "750 GB" hard drive? And in either case, how do you explain the OS's display of "698 GB"? Jeh (talk) 03:40, 6 April 2008 (UTC)
Disagree - way too much clutter and inconvenience. Definitely a good idea in some contexts, but definitely not "the best" in all. — Omegatron 06:27, 6 April 2008 (UTC)

The best method to disambiguate “megabyte” is to precede it with an adjective like "decimal" or "binary" ("binary megabyte" or "decimal megabyte")

Disagree Thunderbird2 (talk) 17:38, 29 March 2008 (UTC)
Disagree – and even if there were a "best method", I'm certain that this wouldn't be it. SamBC(talk) 19:51, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree Fnag aton 16:29, 1 April 2008 (UTC)
Ambivalent If this was also associated with no further use of “mebibyte”, then it might be acceptable. However, advertisements, the popular press, etc. don’t perceive the need to disambiguate this way. So this proposal too would have the effect of leading rather than following common industry practices. In other words, this would have the effect of attempting to solve a “problem” that the industry doesn’t seem to even recognize is a problem. Greg L (my talk) 03:27, 31 March 2008 (UTC)
Disagree Not even acceptable. This approach is only meaningful to those who are already familiar with the issue; also per Greg L's reasons.--agr (talk) 20:20, 2 April 2008 (UTC)
Ambivalent this might work in some cases. Jeh (talk) 03:41, 6 April 2008 (UTC)
Disagree — Omegatron 06:28, 6 April 2008 (UTC)

An acceptable method to disambiguate “megabyte” is to precede it with an adjective like "decimal" or "binary" ("decimal megabyte" or "binary megabyte")

Agree Thunderbird2 (talk) 17:36, 29 March 2008 (UTC)
Undecided a user would have to already know about the difference, and it would be harder to wikilink than using IEC prefixes. SamBC(talk) 19:54, 29 March 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree It is inventing teminology when it is not needed and usually not from the sources relevant to the article. Fnag aton 10:00, 30 March 2008 (UTC)
Ambivalent If this was also associated with no further use of “mebibyte”, then it might be acceptable. However, advertisements, the popular press, etc. don’t perceive the need to disambiguate this way. So this proposal too would have the effect of leading rather than following common industry practices. In other words, this would have the effect of attempting to solve a “problem” that the industry doesn’t seem to even recognize is a problem. Greg L (my talk) 04:26, 31 March 2008 (UTC)
Undecided The arguments presented against (all of them so far) are good ones, so I am withdrawing my 'agree' vote to give me time to think. I'm reluctant to disagree though until an acceptable compromise is found. Thunderbird2 (talk) 17:46, 31 March 2008 (UTC)
Disagree--agr (talk) 20:21, 2 April 2008 (UTC)
Ambivalent this might be acceptable in some cases. Jeh (talk) 03:42, 6 April 2008 (UTC)
Depends on context. — Omegatron 06:29, 6 April 2008 (UTC)

Specifying an exact number of bytes is an acceptable method to disambiguate “megabyte”

Agree Thunderbird2 (talk) 11:36, 29 March 2008 (UTC)
Agree (weakly) SamBC(talk) 13:05, 29 March 2008 (UTC)
Agree somewhat, but certainly not a first choice. Perhaps in a footnote exact numbers could be specified, but in the main body of an article it would be terribly awkward. Seraphimblade ^{Talk to me} 14:54, 29 March 2008 (UTC)
Agree A lot less awkward than using virtually unknwon IEC prefixes. Fnag aton 16:06, 29 March 2008 (UTC)
Agree, although it looks stupid to use an abbreviation and then add the expansion. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)

Comment It seems everyone agrees on this. Why don't you take advantage of this consensus? I expect it's unlikely that everyone would agree on a single best method, so rooting for an acceptable method is all I'd hope for. As nobody here seems to be willing to strictly conform to IEC 60027-2, I'd scrap the new binary prefixes completely for now, outside of articles like "Binary Prefix". Using KB/MB/GB with different meanings with additionally disambiguation (either once per article or once per value) is, at least, still better than relying on common sense, context or whatever isn't explicitly mentioned in an article. --217.87.70.242 (talk) 22:02, 30 March 2008 (UTC)
Compromise, an excellent sensible suggestion. This means not everyone gets their first choice, but of course that is the nature of compromise. Adopting this suggestion of course means that IEC prefixes won't be used in the majority of articles. Fnag aton 00:45, 31 March 2008 (UTC)

Comment: "Scrap the new binary prefixes completely for now" is hardly a compromise. Jeh (talk) 00:14, 5 April 2008 (UTC)
Reply: Not well worded, indeed. What was meant was refraining from retrofitting KiB/MiB/GiB into articles for the purpose of disambiguation. Scrapping them wasn't meant to imply prohibiting them. If some author writes an article which conforms to IEC 60027-2, he should be allowed to. Likewise, if someone writes an article in conformance to the JEDEC standard, that should be fine too. Disambiguation of every single value can hardly be desired. That's why I think each article has to be consistent with a single standard. I believe most articles use the units and prefixes only with a single meaning. It should be clearly stated which one that is even if "obvious" to an expert. Fewer articles mix different meanings e.g., those which mention RAM and HDDs. Use of IEC 60027-2 would be ideal for those but if people want their MiBs without accepting decimal MBs, that's no valid approach. It seems the only commonly acceptable approach is using exact byte counts for disambiguation in such cases. It's a compromise in so far that requirement for disambiguation is acknowledged i.e., expert knowledge isn't taken for granted but it's accepted that before IEC 60027-2 becomes in wide use outside of Wikipedia, IEC binary prefixes aren't the tool of choice for disambiguation. --217.87.56.117 (talk) 15:00, 5 April 2008 (UTC)
Agree To 217.87: Dude! Yes! Greg L (my talk) 02:49, 31 March 2008 (UTC)
Agree I'd like something less obtrusive tho.--agr (talk)
~~Disagree~~ Sometimes. We usually don't need to see "512 MB (536,870,912 bytes)". What is important about the disambiguation is not the exact number of 536,870,912, but that it is significantly larger than 512,000,000. Constantly specifying an exact number of bytes will look even more pretentious and pedantic when the exact number is e.g. 512,000,000. However, I don't think it can be denied that there are occasions when disambiguation to an exact number of bytes is not only acceptable but desirable. Jeh (talk) 03:44, 6 April 2008 (UTC)
Agree - Acceptable in some contexts, but should not be used everywhere. 217.87.70.242 seems to have missed the general agreement here. — Omegatron 06:32, 6 April 2008 (UTC)

The best method to disambiguate “megabyte” is to specify an exact number of mebibytes

Disagree Thunderbird2 (talk) 16:43, 28 March 2008 (UTC)
Disagree Fnag aton 16:48, 28 March 2008 (UTC)
Disagree This is an acceptable method. Tom94022 (talk) 22:44, 28 March 2008 (UTC)
Disagree Continued use of “mebibyte” as a ‘disambiguation’ only requires more reading. Greg L (my talk) 23:02, 28 March 2008 (UTC)
Disagree Who'd say "2 megabytes (2*1000000/1048576) mebibytes"? LeadSongDog (talk) 00:38, 29 March 2008 (UTC)
Disagree, even if you mean an approximate number, because there is no best
Agreed in some cases, where the binary values would be an exact, round number ("a flash drive with a capacity of 512 MiB"), absolutely. On the other hand, they would be terribly awkward in cases such as hard drive capacity, where the decimal values are the ones with the exact, round number ("a hard drive with a capacity of 200 GB"). However, decimal prefixes should be used only to represent decimal values, never binary ones. Seraphimblade ^{Talk to me} 14:57, 29 March 2008 (UTC)
Partially agree, it can be the best one. — Christoph Päper 16:41, 29 March 2008 (UTC)
Neutral Woodstone (talk) 07:11, 30 March 2008 (UTC)
What? — Omegatron 06:34, 6 April 2008 (UTC)

Specifying an exact number of mebibytes is an acceptable method to disambiguate “megabyte”

Agree Thunderbird2 (talk) 11:36, 29 March 2008 (UTC)
Disagree it would look very messy. Approximate number of mebibytes, maybe... SamBC(talk) 13:08, 29 March 2008 (UTC)
Comment It doesn't look messy when used for RAM, as in "Model X has 256 MB (256 MiB) of RAM". Thunderbird2 (talk) 14:42, 29 March 2008 (UTC)
Comment: Thunderbird2, this would be counter-productive and may eventually even cause MiB to become ambiguous as it suggests equivalence. I don't think it's clear whether people who accept "Model X has 268 MB (256 MiB) RAM", would also accept your example. Handling of citations may require different rules but if it's absolutely unacceptable to translate a quote, there should be a proper unambiguous footnote, instead of a potentially misleading insertion. I think some question to vote on is still missing: "Terms like Megabyte should be used consistently among all articles". Your example suggests that you are fine with continued dual meanings. --217.87.66.130 (talk) 15:57, 29 March 2008 (UTC)
- Reply: I don't like the dual meaning of MB, but as of 2008 it is a fact of life. I don't like it that manufacturers of hard drives are punished for standing up for the original definition of of "megabyte", but that too, sadly, is a fact of life. I don't wish to give the impression that MB and MiB are synonymous though. Feel free to add a new thesis if you think that might help identify an area of consensus. Thunderbird2 (talk) 16:15, 29 March 2008 (UTC)
Comment, as above, the use of whichever value produces a good round number is encouraged, but decimal prefixes should never be used in a binary sense. Seraphimblade ^{Talk to me} 14:58, 29 March 2008 (UTC)
Disagree Using virtually unknown units causes more confusion. Fnag aton 16:04, 29 March 2008 (UTC)
Disagree What Fnagaton said. Greg L (my talk) 03:20, 31 March 2008 (UTC)
Disagree, either “256 MB” or “256 MiB”, maybe “256 MB (binary)” or “256 MB₂” but rather not “256 MB (MiB)” and certainly not “256 MB (256 MiB)” or “256 MiB (268 MB)”. — Christoph Päper 16:41, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Limited agree I would find this acceptable for the case where MB is use in the binary sense and if MiB is linked, e.g. 128 MB (MiB). There is no perfect answer and this might be a compromise. --agr (talk) 12:40, 3 April 2008 (UTC)
Agree specifically with the MB (MiB) construct, it's short, unambiguous and those readers who follow the link will learn something! Tom94022 (talk) 16:03, 3 April 2008 (UTC)

Comment' That's an absolutely horrible construct and it's confusing as hell. You can't have IEC prefixes without adopting IEC 60027-2 completely. If this had been intented or desired, there would be something like "MdB" in the standard. If you use MiB, your MB must denote one million. --217.87.120.94 (talk) 16:44, 3 April 2008 (UTC)

Disagree 217.87 is right. This is messy. DavidPaulHamilton (talk) 22:10, 3 April 2008 (UTC)

Specifying an approximate number of mebibytes – e.g. 64MB (61MiB) – is an acceptable method to disambiguate "megabyte"

Agree SamBC(talk) 13:11, 29 March 2008 (UTC)
Agree Thunderbird2 (talk) 14:25, 29 March 2008 (UTC)
Agree for some uses, necessity here would probably best be determined case by case. Seraphimblade ^{Talk to me} 14:59, 29 March 2008 (UTC) (To add to above: Generally, I would likely not agree with this, especially when a round number of megabytes would be an exact value, i.e., hard drive capacities. In such a case, the measurement really is in decimal, so no disambiguation is necessary.) Seraphimblade ^{Talk to me} 19:40, 30 March 2008 (UTC)
Disagree Using virtually unknown units causes more confusion. Fnag aton 16:03, 29 March 2008 (UTC)
Agree, although such instances should be seldom. — Christoph Päper 16:41, 29 March 2008 (UTC)

Question: Crissov, could you explain why you agree with this but disagree with the previous statement? As I understand, the example given in the statement is what you considered inacceptable in the previous case. --217.87.66.130 (talk) 18:21, 29 March 2008 (UTC)

Strongly Disagree Decimal MB should rarely if ever be disambiguated to MiB!!!! The only place I can think this appropriate would be where it is necessary to explain the apparent difference, say between an OS MB and an HDD package MB, but in that case it is probably better to disambiguate MB to common decimal units without prefixes. If this said:

Specifying the same number of mebibytes – e.g. 64MB (64MiB) – is one acceptable method to disambiguate "MB"
[Note the Wikilink in MiB]

I would strong support this. Tom94022 (talk) 19:12, 29 March 2008 (UTC)

Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree—Not widely enough known. TONY (talk) 07:42, 30 March 2008 (UTC)
Disagree What Tony said. Greg L (my talk) 02:52, 31 March 2008 (UTC)
Strongly disagree Totally nonstandard and opaque to non-geeks.--agr (talk) 20:25, 2 April 2008 (UTC)
Comment If you added "agree" then why do you support using a term that does not help to clarify the exact number of bytes? If as some of you were claiming the number of bytes is really important then using a term to round up or down is contradictory. Fnag aton 10:03, 30 March 2008 (UTC)
- Reply It is good practice to round any measurement or specification to a suitable number of significant figures, reflecting the precision of that measurement or specification. There is no conflict between this good practice and the use of unambiguous units. Thunderbird2 (talk) 10:24, 30 March 2008 (UTC)
  - While that might be true for lengths and weights the same it not true for quantities of bytes. Unless you are now admitting that actually the number of bytes is not important? If you are then logically that then means you do not need to disambiguate the term at all since it is safe to leave it as only an approximation to convey to approximate sizes that are being talked about. So which is it? Is it really important to know the number of bytes? Or is the exact number of bytes less important? Fnag aton 10:42, 30 March 2008 (UTC)
    - I'm afraid I don't understand your point. The only difference I can see is that the number of bytes is a discrete variable, whereas lengths and weights are continuous. I don't recall ever having said that the exact number of bytes is important, although it may be for some purposes. What I have said on a number of occasions is that specifying the exact number of bytes is an acceptable way to disambiguate. The problem with the megabyte is not that it is imprecise, but that it is ambiguous. Thunderbird2 (talk) 11:01, 30 March 2008 (UTC)
      - You understand the difference between mathematical constructs and lengths/weights, yes? The former has a meaning which is concrete and the later is a form chosen from some arbitrary physical form. So by your own admission disambiguation in this case is not to increase accuracy? (This is because you do not increase accuracy by having an approximate number of disambiguation units.) Fnag aton 11:20, 30 March 2008 (UTC)
        The purpose of disambiguation is to decrease ambiguity. Thunderbird2 (talk) 12:13, 30 March 2008 (UTC)
        But you are making it more ambiguous by being less exact, remember this number is not some arbitrary length or weight it is a number that comes from a mathematical construct. So since you are advocating a method which introduces another ambiguous (less exact) value for what you are disambiguating why are you doing it. Is it to make the value better understood? Fnag aton 12:32, 30 March 2008 (UTC)
        The point is that 61 MiB is less ambiguous than 64 MB. Thunderbird2 (talk) 12:37, 30 March 2008 (UTC)
        That is illogical because it's still ambiguous since it is not exact. So again I ask, why are you disambiguating one value with an ambiguous value? As you said before it's not because you think the number of bytes is important, so why is it? Fnag aton 12:54, 30 March 2008 (UTC)
        
        You could always make it "~61 MiB" SamBC(talk) 12:57, 30 March 2008 (UTC)
        Which only goes to show that disambiguation term is less exact and is therefore ambiguous. Ignore the whole IEC prefix thing for the moment. In mathematics what is the most exact way (least ambiguous way) to state 16,384 without writing the actual number? Is it: a) 1.64x10⁴ or is it b) 2¹⁴ Fnag aton 13:09, 30 March 2008 (UTC)
        You know as well as I do that the one of those two numbers is exact and the other is not. And you also know perfectly well which one it is, so I wonder why you ask the question. Nevertheless, for the sake of argument, and because I am curious where this might lead, I offer my answer: (b). (~~By the way, (a) is not even close, but I am assuming you meant 1.64e4~~) Thunderbird2 (talk) 13:58, 30 March 2008 (UTC)
        
        Oops yes I made a typo with the powers, should have been 4. Fnag aton 14:14, 30 March 2008 (UTC)
        
        Anyway, back to the point. I take it from your other comments on this page, not only in this section, that you want to see biased POV removed from the guideline, is that correct? Fnag aton 14:20, 30 March 2008 (UTC)
        That depends on exactly what you mean. The whole point of the MOS is to standardise on one particular style. It cannot fulfil that purpose unless it contains guidelines, which are POV by nature. What we should strive for is a balanced POV with a wide consensus. Does that answer your question? Thunderbird2 (talk) 19:06, 30 March 2008 (UTC)
        Then we must approach this from a purely logical argument, that is to say we must remove personal bias. You actually made the point above (12:37, 30 March 2) but didn't continue the reasoning to its logical conclusion. ;) The point, as you put it, is that you contend you are choosing the "less ambiguous" option, but actually you're not. The less ambiguous more exact and explicit option is option (b), you agree with this point, I know you do because you said so. Therefore you should write that this option of using MiB is unacceptable and therefore disagree with it because a better less ambiguous option exists. Fnag aton 22:32, 30 March 2008 (UTC)
        I don't see how you reach that conclusion, and disagree with most of your last response. Let's just agree to differ. Thunderbird2 (talk) 17:09, 31 March 2008 (UTC)
        Fnagaton: the exact number of bytes, to the last byte, is not often important. After all, even a (decimal) "750GB" hard drive is not exactly 750,000,000,000 bytes. But the difference between 64×2²⁰ (67,108,864) and 64×10⁶ (64,000,000) is most certainly significant. Note that this difference constitutes a change in the quoted (i.e. significant) figures. Or if that's not enough, just ask the plaintiffs in that recent "hard drive size" lawsuit. So, yes, disambiguation via "MiB" adds accuracy. It adds even more accuracy where the result is an exact number of MiB but was an approximate number of (decimal) MB. Jeh (talk) 09:18, 5 April 2008 (UTC)
        You are wrong for the rasons I have posted above. A summary of those reasons follows: If as you claim the exact number of bytes is not important then it is not important to disambiguate in the first place. So that means not adding IEC prefixes to occurences of KB, MB, GB etc. Since it more ambiguous to use approximate numbers of mebibytes compared to using an exact number of bytes then chosing to support this option is illogical. Unless you are now admitting that you want biased POV in the guideline? Are you? Fnag aton 09:26, 5 April 2008 (UTC)
        Actually if you look at the votes you'll see I have NOT yet expressed support for this option. I was merely adding to the comment thread here (which maybe should have been in the "should we disambiguate MB at all" question, not this one which addresses a specific method). To address your point: You claim "If as you claim the exact number of bytes is not important then it is not important to disambiguate in the first place." Very well, that is your opinion, often expressed. I will now show you why I disagree with your opinion. It is a simple matter of the degree of error, or ambiguity. Consider a "750 GB" hard drive. My position is that for most purposes of discussion in these contexts (e.g. hard drive sizes) "750GB" is sufficiently accurate. In fact the actual useable size of one of my "750GB" hard drives is 750,153,363,456 bytes. (Fact, hard drives (not even old small ones) don't come in whole numbers of MB or GB.) This is a discrepancy of 0.02%. This is a tolerable error for most purposes, especially since not every "750GB" hard drive will have this exact capacity anyway! But e.g. Windows reports this as only "698 GB". Since the OS means 698 GiB, i.e. 698×2³⁰, that's still the right number (to within three digits anyway), but of course it looks wildly wrong to the user. (The OS also reports the 750,153,363,456 figure, but only in the drive properties dialog.) The difference between 750 (decimal) GB and 698 (decimal) GB (i.e. approximately the capacity the user thinks is missing) amounts to a bit over 6.8%. Actually, maybe the user thinks that 750GiB - 698GiB, about 6.9%, are missing, but anyway... my point is this: The fact that we can (for most purposes) tolerate an error of 0.02% in the "750GB" figure (vs. actual drive cap of 750,153,353,456 bytes) does not prove that we can tolerate an uncertainty, or ambiguity, of 6.8 or 6.9%! So, no, choosing to support this option (if I so choose) does not seem illogical to me at all. By contrast, your position seems to me to be illogical, wildly illogical in fact. Perhaps you can explain why you think that logically, 0.02 is the same as, or even within an same order of magnitude of, 6.8? As for your claim of "biased POV", I don't see where it recommends or prefers the use of binary prefixes (which would be bias in favor), only that it acknowledges that they are a standard and may be used. I don't see how this constitutes bias, so I reject your loaded question that I am supporting a "biased POV". Jeh (talk) 13:19, 5 April 2008 (UTC)
        Which can be expressed in a more precise way rather than resorting to using IEC prefixes with ambiguous rounded numbers. Using the binary prefixes to disambiguate what would be in context a decimal value makes little sense. It is therefore illogical to support this option. Fnag aton 13:30, 5 April 2008 (UTC)
(outdent) One, the additional precision isn't necessary (and furthermore would be very visually intrusive) for most uses. In the case of the hard drive it might even be inaccurate, even while implying a high degree of accuracy, because different "750GB" drives have different exact capacities. But "750GB" is what will appear on the hard drive box, and on most driver makers' labels. Two, "disambiguating" to 750,000,000,000, or even 750,153,363,456 (not exactly a whole number of GB!) won't explain why the user sees "698 GB" either. It seems to me that when we're talking about a "750 GB" hard drive we have to say "750 GB", since that's what the box says, and that "698" number (that is, "specify an approximate number of mebibytes", as the poll question asks) ought to appear as well since that's what the users of hundreds of millions of computers will be seeing in the "real world." Most of them won't look at the drive Properties dialog and find the exact number of bytes, either. Whether the "698" should appear as "698 GiB" or "reported by most OSs as 698 GB" is another question, and one not addressed by this poll question. Aside: See, that's the whole problem with the real world: It's ambiguous and isn't self-consistent, so if WP tries to be consistent with the real world, WP ends up being ambiguous and non-self-consistent too. In my opinion that's bad, and moreover it's bad enough to override the issue of WP user unfamiliarity with binary prefixes. Jeh (talk) 13:48, 5 April 2008 (UTC)
- 1) The precision is important in computing. It is also less intrusive than adding virtually unknown prefixes to articles that do not reference them. 2) Adding a footnotes explains to the reader better than adding virtually unused prefixes to the article. It is possible to be consistent without adding IEC prefixes and by doing so it makes the numbers completely exact, which is better in two ways. So it is still illogical to support IEC prefixes. Fnag aton 14:59, 5 April 2008 (UTC)
  - 1) The precision to the last byte is only important for some uses. But the industry has clearly decided that it is not important always. Otherwise, why does the hard drive box say "750 GB" and the operating system "698 GB", rather than writing out the entire number? If the precision to the last byte is always important in computing, why does Windows Explorer (in details view) report a text file with one character in it as 1 KB (KiB)? It is commonly mentioned here that "the IEC prefixes are not much used in the real world," but in the real world file and hard drive sizes are not commonly expressed in exact numbers of bytes, either. 1a) As for "exact", as I have explained at least twice, it's usually going to be impossible to be "exact" to the last byte for hard drive sizes. If you write the drive size as, say, "750 GB (750,000,000,000 bytes)" you are correctly (to the last byte!) disambiguating the SI sense of GB, but you are not being exact as to the hard drive size. And if you write out "750 GB (750,153,363,456 bytes)" you are a) further confusing the meaning of "GB" by implying that "GB" means 1.000204484608, and b) you are being exact only for one particular model of a drive of that size... It might even be exact only for a particular revision of that model! So the additional digits, even while seeming to provide to-the-last-byte accuracy, really are just providing "noise". That's why the whole concept of significant figures exists. 2) Yes, it is possible to disambiguate without using the IEC prefixes. That doesn't prove that the IEC prefixes should not be used for this purpose, and says nothing at all about their use for other purposes. And, most important it seems to me, if you don't use the binary prefix (or at least a power-of-two divisor) you will not be mentioning the "698 GB" most operating systems will display... let alone explaining it. This, to me, is an extremely significant omission, but one that is unavoidable if "MB" is disambiguated only by writing out the entire number. I will also note that disambiguating RAM sizes by writing out the entire number isn't much better. If you write "512 MB (536,870,912 bytes)" you'll get people saying "so it's really 536 MB? Why does my operating system show only 512?" Jeh (talk) 03:16, 6 April 2008 (UTC)
Depends on context. If it's 64 MiB what's wrong with saying "64 MiB"? — Omegatron 06:37, 6 April 2008 (UTC)

The best method to disambiguate “megabyte” is by consensus on an article by article basis

Agree I see no practical alternative. Thunderbird2 (talk) 10:00, 29 March 2008 (UTC)
Disagree We tried that before, some users (Sarenne) would take their POV to push IEC prefixes to each article creating even more trouble. Fnag aton 16:50, 28 March 2008 (UTC)
Comment I have seen this work in practice. If edits are made without consensus they can be reverted. Where is the problem? Thunderbird2 (talk) 22:36, 28 March 2008 (UTC)
Disagree As this leads to continued use of “mebibyte” as a ‘disambiguation’ that only require more reading. Greg L (my talk) 23:02, 28 March 2008 (UTC)
Disagree Just ducks the issue LeadSongDog (talk) 00:42, 29 March 2008 (UTC)
Undecided, although it's worth pointing out that this is effectively what we're left with "by default" if we can't come to consensus for the guideline; it would be bad to have a guideline indicating a consensus where there demonstrably isn't one. SamBC(talk) 13:13, 29 March 2008 (UTC)
Comment The guideline should not state there is a consensus where there isn't one, but there is a (small) number of points for which there is a clear consensus. These points can be reflected in the guideline. Thunderbird2 (talk)
Disagree, in terms of units of measurement, we need a clear, consistent policy for usage which applies consistently on a sitewide basis. Haphazard "consensus" by a few would-be owners of articles is not the way to go; that would create more ambiguity, not less. Seraphimblade ^{Talk to me} 15:01, 29 March 2008 (UTC)
Disagree, that’s what we have the MoS for. — Christoph Päper 16:41, 29 March 2008 (UTC)
Neutral Woodstone (talk) 07:11, 30 March 2008 (UTC)
Comment I tend to agree with LeadSongDog and Crissov: this can only be an interim solution. In the long term we can hope for something better, but in the meantime I maintain this is the best we have. Thunderbird2 (talk) 18:13, 31 March 2008 (UTC)
Comment A consensus guideline would certainly be preferable, if that can be achieved. --agr (talk) 20:28, 2 April 2008 (UTC)
Comment I would agree only if a consensus guideline cannot be crafted. Jeh (talk) 09:08, 5 April 2008 (UTC)
Disagree - it's not the best method; it's just the most workable method. Nothing else can be reached with the polarized opinions present. — Omegatron 06:39, 6 April 2008 (UTC)

"Megabyte" should be consistently used to denote 1 million bytes in all articles

Disagree, that would be very difficult to do unless we can be completely sure of the intended meaning in every source, or reject any source that we couldn't be sure of in that way. SamBC(talk) 20:02, 29 March 2008 (UTC)
Comment, "should" isn't "must", so exceptions are allowed e.g., exact citations ("640 kilobyte ought to be...") or product names, but it's not meant to be as lax as "if it's about RAM,,, if it's about HDD capacity". This statement explicitly does not mention whether MB has to be disambiguated and if yes, in which way. Sambc, I'd say if it's not possible to figure out how a source uses these units, the editor should either mention this or the source shouldn't be used at all. --217.87.66.130 (talk) 20:43, 29 March 2008 (UTC)
Comment The trouble with this is that, whether we like it or not, the megabyte is an ambiguous unit. (That's one of the few points that gains 100% consensus here). I don't think it is the role of Wikipedia to go around redefining words. Thunderbird2 (talk) 21:57, 29 March 2008 (UTC)

Reply Self-standing MB is definitely ambiguous. As an example, it would be possible to add a disclaimer to each affected article, stating that Megabyte is used in accordance with IEC 60027-2 linked to an explanatory article. Just linking Megabyte might be insufficient if there are no inline disambiguations or similar. --217.87.66.130 (talk) 23:53, 29 March 2008 (UTC)

Disagree. In the real world 1 MB can mean both 1,000 KB and 1,024 KB. It is not the job of an encyclopaedia to define it; it should state, with suitable and sufficient references, what definitions are used.Pyrotec (talk) 22:03, 29 March 2008 (UTC)

Comment If you use 1 MB to denote 1,000,000 Byte, that's not a new definition at all. It is one of the accepted definitions. If you already see the need to state which definition is used, wouldn't you agree that it's most sensible to use the same definition among all articles? --217.87.66.130 (talk) 23:53, 29 March 2008 (UTC)

Disagree I thought long and hard about this. An encyclopaedia should certainly aspire to use units consistently, but the megabyte, like the poor old ton, is a lost cause. Thunderbird2 (talk) 23:05, 29 March 2008 (UTC)

Comment If you consider the Megabyte a lost case, would you consider Megaoctet as alternative? There are certainly a very few cases where it cannot be used because the bytes have more or less than 8 bits, in all other cases, byte and octet are synonymous. My standpoint is, cherry-picking KiB/MiB/GiB from IEC 60027-2 because they are not conflicting, is fruitless. The binary prefixes are a compromise for certain contexts like those represented by JEDEC. But(!) the important part of this standard is restoring (or establishing if you will) the original meanings of the SI prefixes for the byte unit. It is more important because it's the much more difficult. Unlike the new prefixes it causes conflicts. In my opinion, it is not a valid approach to use a part of the standard and violate it at the same time, certainly not in the same article. If that's the idea, even the status quo is better because we don't need redundant prefixes just to prove we could do better. That would, in fact, border on space-cadetism. --217.87.66.130 (talk) 23:53, 29 March 2008 (UTC)

Disagree, but usage should always be disambiguated Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree The necessary flip side of this is that mebibyte is continued to be used to exclusively denote binary math. As this is a poorly recognized term for the general reader, it is unsuitable for use in an encyclopedia. Greg L (my talk) 02:56, 31 March 2008 (UTC)
Disagree Wikipedia is a tertiary source. We should follow the lead of the secondary sources, which for the most part use megabyte both ways and add notes as needed.--agr (talk) 20:30, 2 April 2008 (UTC)
Agree - Ideally, at least, we'd follow the standards. Unworkable, though. — Omegatron 06:41, 6 April 2008 (UTC)

"Megabyte"/MB is widely used in non-scientific/non-specialist literature (magazines, advertisements, product specifications) to mean 2²⁰ bytes

Agree SamBC(talk) 20:02, 29 March 2008 (UTC)
Agree Thunderbird2 (talk) 21:59, 29 March 2008 (UTC)
Agree; but it is also used as 10³ x 2¹⁰ bytes.Pyrotec (talk) 22:08, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree on technical merits, sure, it is, but then "cool" or "hot" is often used in such to mean "good" or "popular", and we would never, despite that, use those terms in such a way. This is a formal reference work, those are not, and so we use more formal terminology than ads and magazines. Seraphimblade ^{Talk to me} 08:17, 30 March 2008 (UTC)
- So are specification sheets (a formal reference), and they generally use MB when they want to say 2²⁰ bytes as well. It's not just adverts and vernacular magazines. SamBC(talk) 13:00, 30 March 2008 (UTC)
  - Seraphimblade when you wrote "We are a formal reference work" you missed out the conclusion of thet statement which should be "We are a formal reference work so we must reference the sources that we use to write our articles". Also as SamBC points out reference sheets from manufacturers use MB etc in the binary sense. Fnag aton 13:13, 30 March 2008 (UTC)
    - And as I have explained many times, measurement conversions are certainly allowable in referencing sources. That's done all the time and is accepted well beyond Wikipedia. The standards for the definitions of the measurements are easily accessible and referenced. Seraphimblade ^{Talk to me} 08:15, 31 March 2008 (UTC)
      - You are still wrong because you forgot one thing, the measurement conversion has to improve the terms being disambiguated and at least have a chance of being well known by the average reader, like the metre for example. The IEC prefixes are virtually unknown and as such pushing for them to be used is against the policy of measurement conversions. This is the way Wikipedia works, it reflects common use. Wikipedia does not promote things from standards organisations when those things are not in common use and not used by the sources relevant to articles. Fnag aton 09:30, 31 March 2008 (UTC)
Agree And in the context of RAM and file size, it is usually understood to mean binary values and typically needs no disambiguation. Greg L (my talk) 02:57, 31 March 2008 (UTC)
Agree MB, megabyte and Mbyte are still the predominate units in scientific journals. (Search on IEEE database for articles published in 2007 and 2008.) -- SWTPC6800 (talk) 03:41, 31 March 2008 (UTC)
Agree--agr (talk) 20:33, 2 April 2008 (UTC)
Agree, but it is also widely used to mean 10⁶ bytes. Jeh (talk) 09:05, 5 April 2008 (UTC)
Agree, but it is also widely used to mean 10⁶ bytes. — Omegatron 06:42, 6 April 2008 (UTC)

"Megabyte"/MB is widely used in non-scientific/non-specialist literature (magazines, advertisements, product specifications) to mean one million bytes

Agree Thunderbird2 (talk) 21:59, 29 March 2008 (UTC)
Agree; but it is also used as 10³ * 2¹⁰ bytes.Pyrotec (talk) 22:11, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree And a disambiguation is typically added to make it clear that decimal value applies (see all Seagate data sheets). Greg L (my talk) 03:00, 31 March 2008 (UTC)
Agree--agr (talk) 20:33, 2 April 2008 (UTC)
Agree, but it is also used to mean both 2²⁰ bytes and (less commonly) 10³ * 2¹⁰ bytes! Jeh (talk) 09:02, 5 April 2008 (UTC)
Agree, but it is also used to mean 2²⁰ bytes. — Omegatron 06:43, 6 April 2008 (UTC)

"Mebibyte"/MiB is used very infrequently in non-scientific/non-specialist literature

Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Agree Thunderbird2 (talk) 12:08, 30 March 2008 (UTC)
Agree Greg L (my talk) 03:01, 31 March 2008 (UTC)
Agree--agr (talk) 20:34, 2 April 2008 (UTC)
Agree Jeh (talk) 09:00, 5 April 2008 (UTC)
Agree SWTPC6800 (talk) 15:49, 5 April 2008 (UTC)
Agree somewhat - "very infrequently" is too extreme. It's used infrequently in non-scientific literature. — Omegatron 06:45, 6 April 2008 (UTC)

"Mebibyte"/MiB is, in fact, widely used in all or most areas of writing and publishing to which it is applicable

Disagree SamBC(talk) 20:02, 29 March 2008 (UTC)
Disagree Here is a blog from the British Computer Society's web pages gives a computer professionals' perspective on the IEC standard [1]. It is not widely known by computer specialists if this blog provides a representative viewpoint.Pyrotec (talk) 22:26, 29 March 2008 (UTC)
- And here is a blog from an IBM employee for contrast: [2] — Omegatron 06:55, 6 April 2008 (UTC)
Disagree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Disagree Thunderbird2 (talk) 12:07, 30 March 2008 (UTC)
Disagree Greg L (my talk) 03:01, 31 March 2008 (UTC)
Disagree--agr (talk) 20:35, 2 April 2008 (UTC)
Disagree , of course. Claiming otherwise would be foolish. Jeh (talk) 08:59, 5 April 2008 (UTC)
Disagree - It could be more widely used. — Omegatron 06:55, 6 April 2008 (UTC)

"Mebibyte"/MiB et al are good units that should be encouraged, and wikipedia should "lead the way" in their use rather than follow the real-world use pattern

Disagree. That is not the job of an encyclopedia. If you look at Encyclopedia, its purpose is to collect knowledge and disseminate it, i.e. information should be provided about Mebibyte/MiB and about Megabyte/MB. It's function is not to promote the use of one set of units in preference to another (OK we have a preference for SI units, but this is not an argument over Imperial versus metric units, it is about counting, e.g. do we count in twos, or tens - or even hexadecimal).Pyrotec (talk) 22:21, 29 March 2008 (UTC)
Disagree. It's not our job to promote anything. It is, however, our job to be clear, precise, and unambiguous. Dpbsmith (talk) 22:47, 29 March 2008 (UTC)
Disagree Thunderbird2 (talk) 22:57, 29 March 2008 (UTC)
Agree Woodstone (talk) 07:11, 30 March 2008 (UTC)
Not that simple, I don't think it's Wikipedia's job to either lead or follow. However, when a formal reference work is presented with one choice of terminology which is ambiguous, and one which is unambiguous, it should choose that which is unambiguous. Formal reference works routinely use terms which may not be immediately understood by the layperson. Seraphimblade ^{Talk to me} 08:19, 30 March 2008 (UTC)
Disagree Regarding Seraphimblade’s statement, using terminology that “is not widely recognized by the typical Wikipedia reader” (a statement he agreed with), has the practical effect—setting aside the issue of whether it is regarded as being important or not—of Wikipedia trying to lead. I don’t think it is wise for any encyclopedia to adopt key vernacular before it is widely recognized by the target audience. Greg L (my talk) 03:11, 31 March 2008 (UTC)
Agree Jeh (talk) 07:53, 31 March 2008 (UTC)
Comment I agree with the first half and disagree with the second. We should adopt best practice which will sometimes mean using these units, but I don't see that as leading the way. That's why I disagree with the thesis. It's a close call though. Basically I agree with Seraphimblade. Thunderbird2 (talk) 17:21, 31 March 2008 (UTC)
Disagree This talk page voted on this proposal in July 2005: "The MoS should encourage the use of the IEC prefixes in all binary-multiple contexts."[3] Trying to "lead the way" can be difficult. -- SWTPC6800 (talk) 23:46, 31 March 2008 (UTC)
Disagree We should not push terminology that has not been widely adopted. See WP:SOAP. We can be unambiguous in other ways.--agr (talk) 20:41, 2 April 2008 (UTC)
Disagree - Standards organizations, open-source software, and computer researchers have already led the way. Using standardized units consistently is appropriate for a general-purpose reference work, and does not "promote" anything. — Omegatron 06:58, 6 April 2008 (UTC)

Articles on technical/specialist subject should use Mebibytes, and more basic or general articles should use Megabytes

Disagree — Omegatron 06:59, 6 April 2008 (UTC)

Headings from the above sections that have complete consensus

Where at least five people have added votes these sections from above with the consensus are listed:

Consensus	Heading
Agree	Specifying an exact number of bytes is an acceptable method to disambiguate “megabyte”
Disagree	"Megabyte" should be consistently used to denote 1 million bytes in all articles
Agree	"Megabyte"/MB is widely used in non-scientific/non-specialist literature (magazines, advertisements, product specifications) to mean 2²⁰ bytes
Disagree	"Mebibyte"/MiB is, in fact, widely used in all or most areas of writing and publishing to which it is applicable

Conclusions of the above votes that can lead to changes in the current guideline

For this question "Specifying an exact number of mebibytes is an acceptable method to disambiguate “megabyte”" there are 4 agree and 4 disagree, no consensus. For the other question "The best method to disambiguate “megabyte” is to specify an exact number of mebibytes" which has 6 disagree and 2 partial agree, near consensus against. For this question "Specifying an approximate number of mebibytes – e.g. 64MB (61MiB) – is an acceptable method to disambiguate "megabyte"" there are 5 agree and 4 disagree, no consensus. The results of these votes show that there is no clear consensus of support for the phrase in the current guideline which says "Use of IEC prefixes is also acceptable for disambiguation (256 KiB).". This is because the results of the votes clearly show there is no consensus for saying "is also acceptable" since the votes which use the word "acceptable" have no consensus. Since it is clear there is no clear consensus of support this also means therefore the whole phrase should be removed. Does anyone disagree with those figures? Fnag aton 23:00, 4 April 2008 (UTC)

I strongly disagree with your conclusion, and I belatedly update the figures, because I was unavailable to add my !votes (been busy offline), and will contribute to the totals in the following way: Q1: 4a vs. 4d + 1pa = NC. Q2: 6d vs. 3pa = NC. Q3: 6a vs. 4d = NC. What you are missing is that the (now more positive) no consensus for question 3 is not "no consensus for the current text", it is "no consensus to change the current text", bolstered very strongly by the fact that there is no consensus in favor of either of the proposed replacements (which are not different enough to matter, really, anyway - you might as well call it Q1+2: 10d vs. 4a + 4pa, which at least in XfD terms is usually close enough to a consensus against them. Please look into WP:CCC more closely. Consensus can change, but there has to be a clear consensus on what new consensus to change it to. We cannot game the system by tweaking how consensus is established and undone for our convenience. — SMcCandlish [talk] [cont] ‹(-¿-)› 00:29, 5 April 2008 (UTC)

I note that the original vote to go with current MOSNUM policy (archive #22 here) was 20 for the policy and 7 against. That wouldn’t qualify either as a Wikipedia-style “consensus” (where it must be nearly unanimous); yet the policy was adopted anyway. In fact, I see that Omegatron at that time advocated slipping in the proposed policy promoting the IEC prefixes without any vote whatsoever and suggested that they all just see how it would became one. Consequently, I see the status quo as being somewhat of a double-standard. I don’t believe the proponents of retaining the current MOSNUM policy on binary prefixes can properly argue that it enjoys the shelter of being grandfathered in and should rightfully be protected by the requirement for a Wikipedia-style consensus to change it since there wasn’t a proper consensus for it to be adopted in the first place.

As Fnagaton pointed out above, if the current policy came up again for consideration, a proper consensus would again be absent if one objectively considers the consequences of the above position statements. It seems the proper thing to do is for MOSNUM to be silent on the issue of binary prefixes. Perhaps this will accelerate progress to a proposal that can achieve a proper consensus. That can only be a good thing as this debate keeps on coming up again and again. This should come as no surprise to anyone as this is simply a natural consequence of the current policy having been improperly adopted in the first place.

I would hope that whatever policy is adopted should start Wikipedia on a path for eventual convergence in how all articles across Wikipedia handle binary prefixes. As it currently stands, the terminology used on different articles means different things. Even Wikipedia isn’t imune from the fact that this is not a good practice. Greg L (my talk) 00:30, 5 April 2008 (UTC)

I'm uncertain that you fully understand how WP consensus actually operates, because you keep speaking in terms of "votes" and "unanimity" neither of which have much at all to do with WP:CONSENSUS. A 20 to 7 head-count, provided that the position of the 20 is reasonable, and the objections of the 7 can be rationally addressed even if not resolved, is usually enough to proceed (if it were not, AfD would rarely delete much of anything). It isn't a vote. If it were a 100 to 1 head count and the 100 were utterly full of crap and making irrational arguments, and the 1 was the only one with a logical position, things should proceed in the direction of the 1. This principle is (again) most clearly visible at AfD (e.g. when a big pile of fanbois want to keep some article on a minor Pokemon character or whatever, and a few people who understand how WP works say "well, no", and the article is deleted because the critics provide policy- and precedent-grounded reasons for deletion, and the keepers provide nothing but "I like it"-style arguments); that this happens is the entire basis of WP:AADD.

The gist is that a finding of WP-style consensus cannot be thwarted with nonsense. Nor can it be thwarted by yelling, by wikilawyering, by attacks, by arguments that are fallacious, by stonewalling (i.e. being unwilling to compromise for the good of the encyclopedia because one's pet peeve is too strong), etc., etc. Numbers matter to an extent, as does the analysis of the merits of the positions of both sides regardless of their number. But the telling factors here are: The current text had way more support. It is stable. It is firmly within the spirit of WP and the MOS in that it declares outright that there is no consensus to prefer one over the other, and thus is doesn't mandate one or or the other. It provides guidance on nothing but how to help the reader by disambiguating and clarifying. And it strives to help editors avoid getting into editwars by suggesting that the symbols not be changed unless the editor is absolutely certain what they are doing. And it's been here a quite a while, minor twiddles notwithstanding. And it is largely being actually followed in real practice at WP. That makes it a consensus that has to be changed by the processes outlined at WP:CCC, and this has clearly not happened yet.

WP's brand of consensus is like 10 friends deciding what to do for the evening. 4 want to go to a movie, 3 want to go to a night club, and 3 want to hit the local dive bar. After discussion 2 want to go to a movie, and 7 want to go clubbing, while 1 is still insistent on the bar. Then 8 are for the club, and the remaining movie guy doesn't want to go alone, so he opts for the bar because he's nervous in clubs. More discussion ensues, and the 8 basically say, "look, we eight are going clubbing, and you can either come with us and try to have a good time, or you can go to your bar have a great time if you're lucky or a lame time because your friends won't be there; c'mon." The hold-outs can either shrug and go along (and probably just get over it pretty quickly) or stick to their guns. If they say, "forget you, we're going to go to the bar", this does not change the fact that there was a general consensus of the "community" of the friends to hit the club; not unanimously, but oh well. They bulk of them will still (presumptively) have a good time, it's really doubtful that any friendships will actually fail to survive the disagreement, the hold-outs may or may not enjoy themselves by keeping themselves out of that mass excursion, and life goes on. Next week, clubbing are on the agenda again, but the hold-outs from last week say, "look, it was the club last week, and you guys said you didn't really have all that great a time, but me and Bob had a blast at the bar, you shoulda been there." And everyone agrees to check it out, except that two of the original movie fans are adamant about the movie, and one is going back to the club because he has a hot date there. Same scenario plays out - the movie fan's won't budge, and the clubber sure as heck won't. Consensus changed, despite the majority being in favor of the club again initially, and unanimity not being reached. The more cohesive mass, if you will, of that micro-community is going to the bar by consensus. WP:CONSENSUS and WP:CCC operate like this. — SMcCandlish [talk] [cont] ‹(-¿-)› 05:39, 5 April 2008 (UTC)

SMcCandlish, the point it is not about what you disagree with. The point is also not about showing consensus to change. The point is not how much you can write about a topic that is not relevant to the actual point. The actual point above shows there is a lack of consensus to support "The Phrase" ("Use of IEC prefixes is also acceptable for disambiguation (256 KiB).") and because there is lack of consensus then that means it should be removed from the guideline. Just so you are absolutely clear, I have demonstrated where there is a current lack of consensus and unless you can show where there is current consensus to keep the phrase then it should to be removed. Fnag aton 07:16, 5 April 2008 (UTC)

SMcCandlish you also wrote "there is no consensus in favor of either of the proposed replacements" which is not true, the section "Specifying an exact number of bytes is an acceptable method to disambiguate “megabyte”" has consensus. It now has one late vote by an individual which I find to be ironic considering your quote from above about "stonewalling". Now then, it is absolutely clear that while there is no consensus of support for The Phrase there is consensus of support to use "an exact number of bytes" "to disambiguate “megabyte”". So the guideline should be updated to reflect that consensus and also updated where there is a lack of consensus. Fnag aton 07:38, 5 April 2008 (UTC)

SMcCandlish I also urge you to keep your replies brief and on topic because otherwise it could be seen as going against your quote "Nor can it be thwarted by yelling, by wikilawyering, by attacks, by arguments that are fallacious". Fnag aton 07:38, 5 April 2008 (UTC)

I strongly oppose such a conclusion. Fnagaton, you are cherry-picking and finessing to support your point of view. Fact is that it is not valid to take the results of a straw poll on a number of relatively independent, but sometimes regrettably intertwined, points and use them as justification for a change in the article when no one has expressed an opinion specifically on that change. Nor is it valid to assume that an apparent lack of consensus (based on the results of said straw poll) for a phrase currently in the article is equivalent to consensus for its removal. I submit that your proposal, while perhaps supported by the straw poll votes you mention, constitutes deprecation of the IEC standard, and the first three poll items clearly show that there is no support for such a move. (jeh)

This post above is an example of the parts of SMcCandlish's quote from above regarding "by yelling" and "by attacks". You see you claim "Fnagaton, you are cherry-picking and finessing to support your point of view" without actually giving any hard evidence. The first three poll items do not show what you claim because none of the first three poll section headings use the word "deprecation", to try to claim otherwise would be you applying your biased POV. This is because what I posted are the facts and what you posted is just bad faith accusation. (fnagaton)

Just because the word "deprecation" does not appear does not mean that is not the view expressed. "Deprecate" means "express strong disapproval of" and the phrases "shall not routinely be used" (first straw poll option), "should be banned" (second and third), and "should be discouraged" (fourth) most certainy constitute expressing disapproval, "strong" in three cases out of four.

So, there's the evidence: It appears to me that you ignored the poll results that didn't fit your conclusion. If you choose to consider that a "bad faith accusation", I don't see how that is supported: The poll results are there; in my view they actively oppose your conclusion; and you ignored them. That's QED in my book. But, see, I don't consider that "bad faith" on your part, merely a mistake. Jeh (talk) 11:36, 5 April 2008 (UTC)

(And if THAT constitutes in your view a personal attack, then what of your oft-repeated "you are wrong" proclamations? At the very least, I must say that that is a style of expression I do not often see on WP.) Jeh (talk) 11:36, 5 April 2008 (UTC)

It doesn't matter that you strongly oppose because what matters is the demonstration of lack of consensus for support of The Phrase. So you are wrong because the facts as presented have nothing to do with what you call "point of view", the numbers speak for themselves. (fnagaton)

"It doesn't matter that (I) strongly oppose"? WOW. Do you, after all, consider your proposal to remove The Phrase a "done deal"? Sorry but you do not get to decide what "matters" and what doesn't, nor do you get to dismiss opposing viewpoints. Jeh (talk) 11:36, 5 April 2008 (UTC)

To address your point: As I stated before, the fact that SOME straw poll results "speak for themselves" in apparent support of removal of The Phrase is not definitive. Even in the absence of any other results, this cannot be construed as direct support for removal of The Phrase, for reasons I have described and which you have yet to counter, let alone effectively rebut. Suppose there is significant opposition to removal of The Phrase here: Are you still going to point to those few straw poll results? Jeh (talk) 11:36, 5 April 2008 (UTC)

Lastly, the vote for the third hybrid propsal clearly shows a majority against using the IEC prefixes as described by The Phrase, so you are also wrong there. Fnag aton 09:04, 5 April 2008 (UTC)

On the other hand, the votes on the first four straw poll items even more clearly show a majority against phraseology that would ban the use of IEC prefixes in most contexts -- and those contexts do include disambiguation. So where does that leave us? Time for more discussions and perhaps other proposals, I think. Jeh (talk) 11:36, 5 April 2008 (UTC)

But, perhaps there IS support for your proposal. Why don't you write up the before and after versions of that section per your proposal, and ask for votes? If the consensus is as you claim it is, what do you have to lose? (jeh)

You are wrong because my edit "23:00, 4 April 2008 (UTC)" is a write up of what is proposed. Fnag aton 09:05, 5 April 2008 (UTC)

What is there to be "wrong" about in what I wrote there? "Perhaps there is support for your proposal." Hm, no, not that. "Why don't you write up your proposal and ask for votes?" Ok, you already described it, but you're not quite there: You haven't asked for votes. Jeh (talk) 11:36, 5 April 2008 (UTC)

And speaking of yelling, you might consider not addressing people in such a superior-sounding tone, particularly your near-constant use of "you are wrong." Personally, I read your "you are wrong" as simply "I disagree", and no one else should put any more stock in that phrase of yours than that. Jeh (talk) 08:31, 5 April 2008 (UTC)

I am not yelling though, I am using the facts the prove you wrong. I am also using an exact tone which counters your bad faith personal attack tone you are using. The fact is you are now trying to use ad hominem instead of tackling the facts of my argument. Fnag aton 09:08, 5 April 2008 (UTC)

I didn't say you were yelling, I said "speaking of". Perhaps I should have said "speaking of your accusation of yelling". You took it upon yourself to criticize (or in your word, "attack") another editor's posting style, but you yourself are exempt from any such criticism? Even when introduced with "you might consider"? Please. Jeh (talk) 11:36, 5 April 2008 (UTC)

GregL, your conclusions are not supported. First off, SMcCandlish has shown that the previous adoption of the phrase did achieve "consensus" by WP's rules. If you read WP:CON you will find that nowhere is it stated that support "must be nearly unanimous", only that everyone agrees to abide by the conclusion. There is an important difference.

Second, GregL, you cannot interpret the results of the straw poll as "support" or "oppose" for the current text, any more than Fnagaton can do so for a change to the text. (But if you insist on doing so, I submit that you are attempting to deprecate the use of the IEC prefixes, and that in the votes on the first few straw poll questions you will find overwhelming opposition to that notion.)

The straw poll consists of a number of questions on "small" points, some overlapping, some with two points wrapped into one question, and (most crucially) each having a different weight in each participant's mind. Consider: When I vote for a candidate that doesn't mean I necessarily agree with all of that candidate's positions on the issues. It means that that candidate is the best compromise as far as I am concerned. Someone reading my answers to a straw poll such as the one here might conclude that I would vote a different way, but they don't know my priorities. It might be that the candidate I vote for agrees with me on only three out of twelve issues, but that those three are the most important to me.

As I suggested to Fnagaton: if you want to find out how people would vote on the current phrasing, if it were proposed now, why not make that a poll item? It's not as if it's difficult.

The straw poll is useful for developing suggested changes. But you can't take its results and then interpret them as support for a change that hasn't been specifically discussed here. Jeh (talk) 08:53, 5 April 2008 (UTC)

You are wrong because the very fact that I created the section proposing this means the change is specifically being discussed here. Fnag aton 09:19, 5 April 2008 (UTC)

I am glad you agree that we are DISCUSSING the change here, and you are not already concluding that support has already expressly been given for your proposed change. Therefore I hope you will set up a vote on exactly your proposed change. Not something "close to", not something "supported by" the straw poll, but exactly your proposed change. If you don't, I will, but my wording might not exactly match what you want. Jeh (talk) 09:45, 5 April 2008 (UTC)

I note another bad faith attempt at misrepresentation because what you are writing has no basis according to the facts above, my first post in this section "23:00, 4 April 2008 (UTC)" opens the matter up for discussion. You will note the language "should be" is not the same as "must be" for example. The change is already explictly stated above by my summary at the start of this section, you may either add your agree or disagree as you see fit in this section and treat it as a poll if you wish. But first you should answer the question put directly to you below in post "09:54, 5 April 2008 (UTC)". Fnag aton 10:27, 5 April 2008 (UTC)

A "bad faith attempt at misrepresentation"? Where? What did I misrepresent? What did I say you said, that you didn't say? I expressed gladness that you consider your proposed change open for discussion. Jeh (talk) 11:36, 5 April 2008 (UTC)

I do note, however, that you concluded that edit by asking only "Does anyone disagree with those figures?" as if to suggest that the only things open for discussion were the specific figures you had quoted. If I was in error in thinking you might have intended to narrow the discussion in that (or any other) way, then again, I am glad. Jeh (talk) 11:36, 5 April 2008 (UTC)

Bad faith misrepresentation by trying to imply I am not discussing this when in actual fact I am. You are in error. Fnag aton 12:19, 5 April 2008 (UTC)

I also have extreme misgivings when you tell me "It doesn't matter that you strongly oppose because what matters is the demonstration of lack of consensus for support of The Phrase." That's not a "representation", "mis-" or otherwise, that's a direct quote from you. Are you in fact agreeing that your proposal (not just a few points in support of it) for removal of The Phrase is completely open for discussion? If so, again, I am glad. Jeh (talk) 11:36, 5 April 2008 (UTC)

That is an irrelvant question because it is a strawman based on you being in error above. Fnag aton 12:19, 5 April 2008 (UTC)

"If you read WP:CON you will find that nowhere is it stated that support "must be nearly unanimous", only that everyone agrees to abide by the conclusion." - I'm glad you wrote that because you're now admitting there doesn't need to be a unamimous conclusion. So looking near the top of the page the results of the third hybrid proposal show a clear majority approving the proposed text. Do you agree to abide by the conclusion? Fnag aton 09:15, 5 April 2008 (UTC)

One, you are being unnecessarily argumentative: I never claimed there needed to be a unanimous conclusion (in fact that was your friend here GregL), so your crowing that I am now "admitting" there doesn't need to be one is completely unfounded. Two, SMcCandlish's points apply: Just because there is a 13 to 9 majority does not mean "the motion carries" here. Particularly since the third proposal clearly constitutes deprecation of the IEC prefixes and the first three -- no, four straw poll questions clearly show opposition to such a move, in fact with a far greater ratio than 13 to 9. Jeh (talk) 09:45, 5 April 2008 (UTC)

I am not "being unnecessarily argumentative" and I note another of your bad faith accusations. (fnagaton)

Would you like to address the fact that you were attributing to me a position I never held? Jeh (talk) 11:36, 5 April 2008 (UTC)

The fact is that you are wrong, this has been demonstrated and you have not refuted my point. Again, the first three (or four) votes do not support what you claim because they are on different issues and "you are cherry-picking and finessing to support your point of view". (fnagaton)

The view the first three or four straw poll items express is on all fours with your proposed removal of "the phrase." And they were all soundly rejected. Jeh (talk) 11:36, 5 April 2008 (UTC)

Yes I just used your own quote, this is to demonstrate exactly how illogical your "argument" is. (fnagaton)

On the contrary. You were using a few poll results to show support for your proposal. I'm adding a few more poll results to show that there are strong dissenting views to your proposal. But I'm not denying that your poll results exist, nor am I denying that they are applicable. Jeh (talk) 11:36, 5 April 2008 (UTC)

The third propsal does not "constitute deprecation of the IEC prefixes" as you claim, in actual fact the third proposal does not support the pushing of the POV that IEC prefixes can be used, that is the key difference. The third proposal reduces the bias in the current guideline by relfecting real world consensus better. (fnagaton)

The status quo is to allow the use of IEC prefixes in many contexts. I don't see that it "pushes" anything, it merely allows their use. The third proposal expressly states that "it is no longer permissible on Wikipedia to routinely use the IEC 60027-2 binary prefixes" except in quoted text, descriptions of the prefixes themselves, etc. That most certainly constitutes deprecation of current practice, and it is most certainly in conflict with the results of the first three or even four straw poll items. Please also note that while a change to the status quo requires consensus, there is no requirement I can find to repeatedly demonstrate consensus for keeping the status quo. Jeh (talk) 11:36, 5 April 2008 (UTC)

Lastly (here), your assertion that it "reduces the bias" is your opinion. It furthermore presumes that bias exists, but I see no bias in merely allowing. Jeh (talk) 11:36, 5 April 2008 (UTC)

Also your statement "a 13 to 9 majority does not mean..." contradicts your earlier post "you will find that nowhere is it stated...". (fnagaton)

Let's complete that quote, it was "you will find that nowhere is it stated that support "must be nearly unanimous", only that everyone agrees to abide by the conclusion." And that's accurate from what I can see in WP:CON. I don't see how this contradicts "Just because there is a 13 to 9 majority does not mean "the motion carries" here." WP:CON is in fact quite clear that results of votes, while useful to determine opinions, do not by themselves constitute consensus. For example, "Wikipedia's decisions are not based on the number of people who showed up and voted a particular way on a particular day; they are based on a system of good reasons." Jeh (talk) 11:36, 5 April 2008 (UTC)

The 13 to 9 majority clearly does show majority support for the proposed text, to try to claim otherwise is illogical. (fnagaton)

I'm not claiming otherwise. I agree that that particular vote clearly does show majority support for the proposed text. However I am claiming that the votes on the first three or four straw poll items show the opposite conclusion. We do therefore not have consensusas defined by WP:CON. Indeed, WP:CON clearly states that "Wikipedia's decisions are not based on the number of people who showed up and voted a particular way on a particular day; they are based on a system of good reasons." Many good reasons have been expressed in opposition to your and GregL's proposals, and until I and several others here are satisfied as to our objections, you cannot claim any sort of consensus, even though you have a majority on one particular vote on one particular proposal. Jeh (talk) 03:29, 6 April 2008 (UTC)

So I have to ask again since you did not give a valid answer last time, do you agree to abide by the conclusion? Fnag aton 09:54, 5 April 2008 (UTC)

You may feel you "have to ask", but I'm not required to give any answer, let alone an answer that you'll consider "valid." I am interested in arriving at a conclusion supported by consensus; take that for what you will. I suspect that any further discussion along this line will be fruitless. Why don't you set up a vote on your specific proposal? Jeh (talk) 11:36, 5 April 2008 (UTC)

I already said you can put your vote in this section if you wish, the exact proposal is in my first post of this section. I'll make it simpler for you, this section is to be used for votes for the proposed change to the existing guideline which is at the start of the section. So to be completely clear, you wrote "only that everyone agrees to abide by the conclusion" and yet when I ask you the question "do you agree to abide by the conclusion?" you refure to answer. Your refusal to answer your own statement is interesting. Fnag aton 12:58, 5 April 2008 (UTC)

Also you wrote "The straw poll consists of a number of questions on "small" points, some overlapping, some with two points wrapped into one question, and (most crucially) each having a different weight in each participant's mind." - Putting it another way this means "even if some people have reservations about some individual parts of the proposal they can still support it as a whole", don't you agree? Fnag aton 10:54, 5 April 2008 (UTC)

Yes, it COULD mean that. But it mightalso mean "som people have enough reservations about some individual parts of the proposal that they cannot support the proposal as a whole." Don't you agree that that's a possible conclusion? Jeh (talk) 11:42, 5 April 2008 (UTC)

That is irrelevant and I'll explain below as to why. Fnag aton 11:46, 5 April 2008 (UTC)

Blimey, someone ought to knock some heads together here. Regardless of who's been more unreasonable, or was unreasonable first, most of the contributors to this subsubsection have been unreasonable now. Calm down, let's try a different approach. (Sambc)

I'm not uncalm. Except when someone tells me my opinions and arguments "do not matter". Jeh (talk) 11:42, 5 April 2008 (UTC)

Sambc you spoiled my fun. :( As I was going to post in reply to Jeh this "can't draw a conclusion from a minority subset of the straw polls" applies to the case for when he tries to take the first three/four polls and try to claim that means IEC prefixes cannot be deprecated and it shows why his argument is illogical and therefore wrong. This is because if you look at the poll results for the third hybrid proposal then there is clear majority support for the proposed wording, regardless of his claim of what the first three/four votes mean. This is of course because the proposed text as a whole outweighs whatever individual concerns might be expressed in the "micro votes" on those individual issues. So Jeh is absolutely sure, his claims that the results of the first three/four micro votes "clearly show opposition to such a move" is completely irrelevant according to his earlier post and now your post. Yes I have again used Jeh's own quotes to refute his later points, again this is to show how illogical his argument actually is. Fnag aton 11:50, 5 April 2008 (UTC)

Except no-one has expressed agreement with your reasoning. In fact, I get the feeling that several people feel it is faulty, and at least one has explained why, and been confronted simply with contradiction from you. You are not the arbiter or logic and reason, in fact no individual is, especially here. Get off the high horse. SamBC(talk) 12:07, 5 April 2008 (UTC)

You are wrong, as shown by this comment "As Fnagaton pointed out above, if the current policy came up again for consideration, a proper consensus would again be absent if one objectively considers the consequences of the above position statements. It seems the proper thing to do is for MOSNUM to be silent on the issue of binary prefixes." from 00:30, 5 April 2008 (UTC). Fnag aton 12:22, 5 April 2008 (UTC)

Being more rational

We can't draw a conclusion from a minority subset of the straw polls; we need to consider them as a whole, possibly discarding the ones that were felt to be faulty as premises. So why don't we do that instead, hmm? SamBC(talk) 11:22, 5 April 2008 (UTC)

Which brings us back to the third hyrbid proposal text. Fnag aton 11:32, 5 April 2008 (UTC)

Ah, it seems to me that it brings us back to refining the straw polls. Jeh (talk) 11:42, 5 April 2008 (UTC)

To Fnagaton: Many editors, including myself and Jeh, have expressed reasoned opposition to the deprecation of IEC units. It is clear to us that you believe that they should ~~not be permitted~~ only be permitted in exceptional circumstances. It should be equally clear to you by now that we believe ~~they should~~ the proposed restrictions are unwarranted. The way towards consensus is to reword the proposal (which contains many good elements) without said deprecation. Thunderbird2 (talk) 11:49, 5 April 2008 (UTC)

By the way "It is clear to us that you believe that they should not be permitted" is not an accurate summary of my position so please do not repeat it. Fnag aton 12:31, 5 April 2008 (UTC)

Even with the strikeouts and replaced text that still doesn't accurately reflect my position. There is nothing about "exceptional circumstances" in my position. My position is accurately summarised here in this diff. You will notice it is much more unbiased than your statement implies, this is because I advocate removing "no consensus for using IEC prefixes" and also removing any hint that one standards body is better or worse than another. This is acomplished by not advocating any specific standards body thus being completely neutral on the matter. Fnag aton 13:38, 5 April 2008 (UTC)

Jeh is not using reasoned opposition though because he fails to apply his own "points" to his own position, this I've shown by using his own quotes to refute his own position. ;) Also he is using too many personal attacks instead of tackling the fatcs of the matter, another reason why he is not using reasoned opposition. The problem is you say deprecation when the actual problem is that the guideline pushes a point of view that IEC can be used which is contrary to real world consensus. Many other editors do not want to see that kind of POV pushing in the guideline.[break]

If Jeh has made those mistakes, then it is as nothing to the extent that you have made them. You argue in circles, tell other people they they are wrong without any qualification, and claim victory using logical paths that no-one else seems to see the validity of. Your conduct is most unbecoming, please take a break. No final decision will be made in the next 24 hours, why not take them off and chill? SamBC(talk) 12:03, 5 April 2008 (UTC)

You are incorrect because I always qualify why someone is wrong, especially with the case of using Jeh's own text to refute his own position. I note your use of ad hominem instead of tackling the real issue. I am chilled, it is you who is using personal attacks and not being civil so I suggest you go off and relax, or you can apologise. I don't mind which. I also do not "argue in circles" as you put it, what I was doing was getting Jeh to say enough so that he would refute his own position. Fnag aton

This means, of course, that unless you compromise on the position of pushing for IEC prefixes to be included in the guideline then no matter how long we debate this it will always come back to that point. There has been compromise from the other side by allowing IEC prefixes in articles where the reliable sources use them, so now it is your turn to compromise. So if you are not going to compromise then the vote on the third hybrid proposal stands, which means it has majority support. Do you agree the third hyrbid proposal has majority support? (You just have to look at the numbers.) Fnag aton 11:54, 5 April 2008 (UTC)

Please do not edit your points after you have posted them; it makes it harder for people to respond. Add another point instead, or use the preview feature until you are happy with what you've written. SamBC(talk) 12:03, 5 April 2008 (UTC)

Until someone responds to something I'll edit it as many times as is needed to make the point as clearly as it is possible to make it in the limited time available. Fnag aton 12:29, 5 April 2008 (UTC)

My point is that it makes it very hard to respond; it took me at least 3 attempts to respond to you there, because you kept changing it and adding points. Fixing typos is one thing, but adding stuff to an existing paragraph is rarely necessary, as a second paragraph can be added instead; thus it isn't worth the difficulty it causes the editors who are trying to discuss things with you. SamBC(talk) 15:08, 5 April 2008 (UTC)

Edit conflicts are a way of life here. ;) Fnag aton 15:10, 5 April 2008 (UTC)

No-one would disagree that it had majority support at the time people were !voting on it. However, there is reason to reject the idea that it has consensus, and there's also ample reason to suppose that it might not have majority support now, as different people are in the discussion, and there has been debate that may have changed minds. SamBC(talk) 12:05, 5 April 2008 (UTC)

We come back to what consensus is again. Fnag aton 12:29, 5 April 2008 (UTC)

Call for arbitration then

All: It seems that SMcCandlish has advanced some well researched reasons quoting existing Wikipedia policy on how the current MOSNUM policy on binary prefixes was properly adopted. This is something Fnagaton and I don’t agree with. This reminds me of negotiations during the Vietnam war when the parties couldn’t even agree to the shape of the negotiating table they were supposed to eventually sit around. The two sides here—myself included—have got to find some common ground. The whole notion of pretending that the current MOSNUM policy on binary prefixes was properly framed and therefore should stay as current policy has the effect of leaving things precisely as they have been since it was first adopted: with nothing but conflict among editors. Further, there are now different Wikipedia articles that follow different methods of denoting computer memory. And the result of that? There is no easy way for readers to discern which convention is used. That’s complete garbage. So…

I hereby make a call that we bring an administrator in to arbitrate on this. So I ask this: who here refuses to even accede to the intervention of arbitrator? Is there anyone here who wants to take the position that the current policy was properly framed, and 2) they are so convinced of their position on this point that they won’t budge and even accede to arbitration? Greg L (my talk) 18:19, 5 April 2008 (UTC)

Okay, I'm not sure what you're suggesting here. There's arbitration, which is a fairly long process not really applicable. Or there's formal and informal mediation, and I can't remember whether they'd accept this question, and in either case wouldn't determine an outcome, just help the parties involved to do so. Administrators have no extra standing to determine cases like this either. SamBC(talk) 18:33, 5 April 2008 (UTC)

It doesn’t have to be these more complex routes. The simplest possible path. We simply agree to have an uninvolved volunteer administrator (that rules out Omegatron) step in and we agree to abide by that individual’s guidance and decisions. Greg L (my talk) 18:53, 5 April 2008 (UTC)
- That could only possibly work if the administrator also had no pre-existing opinion on the matter. It is my feeling that any administrator (or any user) who'd be good for the job would, for the same reasons they'd be good, not be prepared to do it. It also falls so far outside normal wikipedia practice as to make me uncomfortable, and I expect others. Your preemptive ad hominem ("so convinced of their position on this point that they won't budge and eve accede to arbitration", as if that's the only reason to object) is also rather disturbing. SamBC(talk) 19:08, 5 April 2008 (UTC)
  - Oh common SamBC, my “preemptive ad hominem” was “rather disturbing” because you took it precisely as I intended: “I double dare you.” Are you going to get all sideways on a little challenge here? It wasn’t even directed so much at you as it was to some others who I anticipate might be more inclined to object to arbitration; you’ve seemed to be more flexible. Greg L (my talk) 19:17, 5 April 2008 (UTC)
    - Actually, that's exactly what I thought you were doing, and that's my problem. You're suggesting that anyone who has any problem with your suggestion is trying to dodge things, and that's a pretty big, non-AGF assumption, as I've tried to illustrate. I am flexible, both in methods and outcomes, but everything has an elastic limit. SamBC(talk) 19:47, 5 April 2008 (UTC)

I would propose we ask User:Theresa knott (contribs). I know she’s good but have no idea if she’d take this roll on. I had a run-in with another administrator who I thought had a “little Napoleon” syndrome problem and power had gone to his head. It didn’t seem that she “intervened” in the literal sense, but she cooled me down and things with the rogue admin seemed to just go better after she made a comment or two for us to both look at. If someone wants to propose another admin, that’s fine; we can contact them both and see if they want to tag team, or accept, or decline, or whatever. The only criteria is they should be 1) an extremely experienced admin, 2) who stays out of computer-related articles and 3) typically has only peripheral involvement with MOSNUM. Greg L (my talk) 19:11, 5 April 2008 (UTC)

A single arbitrator is probably a bad idea - any one person can make a misteak.

This does raise the question of what to arbitrate or mediate, I see two possiblities

1. There appears to be agreement that the original policy was achieved by consensus but there is now a question over the meaning of the sentence:

"There is no consensus to use the newer IEC-recommended prefixes in Wikipedia articles to represent binary units"

Some editors seem to think this means the consensus is to not use IEC-recommended prefixes while other editors seem to think the consensus is to allow editors the choice of using IEC-recommended prefixes where the editor views them appropriate. I suggest starting with informal mediation is appropriate. If necessary we can continue the resolution process thru arbitration. It the question that started this whole discussion and I think can have a yes or no answer.

2. Some editors want to rewrite the entire MOSNUM on binary prefixes. I doubt if any mediator or arbitration panel wants to go the the detail hereinabove or below. Perhaps if we get an answer to the above question thru mediation or, if necessary, arbitration, then the folks that disagree with the answer can craft alternative MOSNUM on binary prefixes and then present that for arbitration by a panel. Tom94022 (talk) 19:37, 5 April 2008 (UTC)

We can’t have a situation where the waring parties here try to negotiate “what’s on the table for discussion” and “what each party is willing to put up to arbitration.” The point of arbitration is break a deadlock and everything has to be on the table. The parties simply have to agree to two points: 1) that there is a significant number of people on both sides of the issue, and 2) that they agree that nothing is “on” or “off” the table when arbitration is called for. If we can’t agree to solve this on our own, then there a number of ways to resolve it in the long run. Proponents of keeping the current policy are completely deluding themselves if they think they are somehow grandfathered in. Wikipedia policy on “consensus” speaks of “consensus minus one”, “consensus minus two or three”. A policy that was adopted with a 20:7 in favor was improperly adopted in the first place and therefore does not enjoy the immunity from change afforded by Wikipedia’s “consensus”. MOSNUM policy is a guideline, for other editors to follow. It isn’t some sort of law that has to be observed like that governing personal attacks. Anyone can edit MOSNUM guidelines; as Omegatron recently demonstrated in the last few weeks. Greg L (my talk) 20:11, 5 April 2008 (UTC)

- GregL, can you please tell me where in WP:CON it says anything about "consensus minus one" or "consensus minus two"? I sure can't find it. (Edit: And neither can Google, at least not on any actual WP policy pages.) I do find on WP:CON an explicit statement that "Wikipedia's decisions are not based on the number of people who showed up and voted a particular way on a particular day." Where exactly do you find these references to e.g. "consensus minus one"? Jeh (talk) 06:07, 6 April 2008 (UTC)

I agree with the call for arbitration and I agree with Greg's choice of uninvolved administrator. The three criteria from Greg (19:11, 5 April 2008 (UTC)) is also very important for the reasons he stated. Also as Greg points out to Tom94022 we cannot have arbitration on one tiny micro issue (Tom's post at 19:37, 5 April 2008 (UTC)) for the same reasons as those posted by myself and SamBC regarding drawing a conclusion from a subset of the straw polls. The issue is neatly summarised by the majority vote in favour of the third hybrid proposal text and it is that vote that is the most recent, extant and conclusive test of consensus. Fnag aton 20:52, 5 April 2008 (UTC)

- "Arbitration" is, according to WP:ARB, a) conducted by the Arbitration Committee, not someone of our choosing, and b) supposed to be the last stage in [[WP:DR|dispute resolution] here. Have you actually read WP:ARB? It seems to me from WP:DR that the next step should be either informal mediation (conducted by the "mediation cabal") or formal mediation. Either way, you don't get to pick the mediator. That said, I think asking for informal mediation would be a good step. But I disagreee regarding the points of dispute. I don't think you can just ask a mediator to come in and "resolve the dispute." We don't all have to agree on all of the points of dispute, but I think it would be worthwhile for each of us to list the points each of us feels are in contention. Jeh (talk) 06:07, 6 April 2008 (UTC)

If 20:9 is not consensus then 13:9 (EDIT: Now 13:10 - jeh) certainly is not any consensus for change! This started as a simple dispute between Fnagaton and Omegatron over the meaning of the one sentence; to me that sounds like a dispute that can and should be resolved thru the Wiki conflict resolution processes. I for one would not agree to a single arbitrator. To expect a group of administrators (or even one) to take the time to go thru all of the above and below and then draft a policy is unrealistic. On the other hand, taking the simple question forward will resolve the issue. If Fnagaton won't start the dispute process on this one issue, I hope Omegatron does. Tom94022 (talk) 00:44, 6 April 2008 (UTC)

This huge discussion certainly isn't about "one sentence". I was going to start dispute resolution against Fnagaton with another editor co-signing in January over his revert warring then, but got sidetracked by real life. I really don't have a lot of patience for this stuff. We obviously need some mediation from third parties who aren't going to have their own opinions on the subject. Biased admins like myself are obviously no help. — Omegatron 07:06, 6 April 2008 (UTC)

Which you should have done and I would have welcomed because dispute resolution would have told you to talk about your changes to MOSNUM and to try to get consensus first and to not try to edit war your changes into MOSNUM guidelines and also you would have been told not to make personal attacks and pathetic threats about blocking. This is documented in the ANI against you and which was counter signed by two other editors. I told you before, do not try to misrepresent the facts because the documented facts point against you since you are in the wrong here. Fnag aton 09:31, 6 April 2008 (UTC)

Anything that breaks this logjam will be a welcome change. It’s clear from the spectacular amount of polling on various points above that we editors here have put in a good-faith effort to understand each others’ position in an effort to try to find some common ground. What other techniques are available to resolve this that we haven’t already tried? I don’t care if it’s “informal” or “formal” or whatever; if it’s new, we should try it and get ourselves of this rut. Greg L (my talk) 01:43, 6 April 2008 (UTC)

Points for mediation

I propose that all who are involved in this dispute should add the questions they would like a mediator to address to this section. I propose that if you don't agree that a question someone else has added here should be addressed by a mediator, simply add whatever points you do want addressed. This list will then be a union (set theory) of all parties' requests for mediation, rather than an intersection.

I would like to strongly discourage discussion of these points in this section as I think it would help things if we just had all of the points listed, as succintly as possible. Of course this is not at all meant to discourage discussion of these points, but I feel it would be better to do that in another section. Jeh (talk) 08:25, 6 April 2008 (UTC)

Question: 1. Whether or not the current MOSNUM policy on binary prefixes, prior to the beginning of the current dispute (need a more precise date here, and a link to MOSNUM of that date), represented consensus at the time according to the guidelines in WP:CON, even though support in the discussion was not unanimous. Jeh (talk) 08:25, 6 April 2008 (UTC)

Question: 2. Whether or not consensus support has been shown for the "third hybrid proposal", and that therefore WP:MOSNUM#Binary_prefixes should be changed to the wording of that proposal. Jeh (talk) 08:25, 6 April 2008 (UTC)

Question: 3. Whether or not the votes on the first four items of the straw poll show consensus that the current MOSNUM policy on binary prefixes should not be changed to discourage or prohibit their use to a greater degree than is currently supported by that policy. Jeh (talk) 08:25, 6 April 2008 (UTC)

Question: 4. Whether or not the votes on the first four items of the straw poll are sufficient to show lack of consensus for any proposal that would discourage or prohibit the use of binary prefixes to a greater degree than is currently supported by MOSNUM. Jeh (talk) 08:29, 6 April 2008 (UTC)

Question: 5. Referring to the previous discussion that concluded with the current wording (need a link here to the archived discussion), whether "There is no consensus to use the newer IEC-recommended prefixes in Wikipedia articles to represent binary units" should be interpreted to mean a) that the consensus is to not use IEC-recommended prefixes, or b) that the consensus was to allow editors the choice of using IEC-recommended prefixes where the editor views them appropriate. Jeh (talk) 08:34, 6 April 2008 (UTC)

I disagree with this for exactly the same reasons as before since you cannot try to use results of individual straw polls or micro issues to try to push through your own point of view. The arbitration must only use the third hybrid proposal text which has been voted on. Fnag aton 09:47, 6 April 2008 (UTC)

Building on common ground

Well, I think there is common ground, and that it’s worth trying to build on it. For a start, we all agree that the megabyte is an ambiguous unit, right? And nearly all of us (9 out of 11) agree that its disambiguation should be encouraged. Perhaps the other two have good reasons for not (yet) agreeing, so let’s ask them.

Fnagaton: why do you disagree with the statement Disambiguation of “megabyte” should be encouraged?
SamBC: why are you undecided about the same statement?
To both: Can you suggest a way of rephrasing the statement in a form that you can agree with?

Thunderbird2 (talk) 21:21, 5 April 2008 (UTC)

- Thank you for reminding me, no we don't all agree on the ambiguous "thing", I have to change my vote because just because something has two definitions it doesn't mean it is ambiguous to everyone (actually I think most people who read the article and need to know about byte sizes will already know the finer details from the context) and because the word ambiguous is being used by some to push their POV that IEC should be used instead. The word "ambigous" has too much POV attached to it to be a useful. As for the encouraging disambiguation 1) It expresses a point of view that is not accurate in all situations and is not relevant to all situations. 2) Advocating/mentioning/encouraging disambiguation of prefixes with IEC prefixes is not acceptable because it is pushing a particular point of view. Fnag aton 22:14, 5 April 2008 (UTC)

- As for wording that I would be happy with, look at the "First draft for consideration". In that proposal you will see no advocacy of either system, you will also not see any unfair/unbalanced deprecation of either system. You agree with those statements, yes? Another example would be "These prefixes (KB/MB/GB etc) are defined with decimal and binary meanings so disambiguation may be needed". Fnag aton 22:21, 5 April 2008 (UTC)

IMO, "without using any other prefix styles" is unfair and unbalanced deprecation of the unambiguous IEC binary system. Tom94022 (talk) 00:31, 6 April 2008 (UTC)

No it's not unfair or unbalanced deprecation of IEC prefixes because the definition of deprecation is "to express earnest disapproval of". Nowhere does my proposal specifically express earnest disapproval of only IEC prefixes. Unless you can prove the contrary then you're going to have to concede that point.

In the current text, specifically mentioning IEC prefixes is to advocate IEC prefixes and that is pushing a biased POV. Putting it another way, specifically mentioning IEC prefixes is actually unfair and unbalanced advocation of IEC prefixes. Pushing a biased POV towards either of the prefixes is unacceptable. Therefore the only choice is to find the middle ground and that means removing any biased POV towards either prefix style. Which is why I proposed the text. Fnag aton 09:43, 6 April 2008 (UTC)