Wikipedia talk:Manual of Style (dates and numbers)/Archive B5

Mediation

Since discussion on this page regarding the prefix issue has failed to reach any consensus, nor to stop revert warring, I think that the next step should be to open a mediation request. Does anyone object to that? *** Crotalus *** 01:21, 10 May 2007 (UTC)

I think that mediation may be the best path. If a real consensus should happen with the current method the mediator could just confirm it. -- SWTPC6800 02:33, 10 May 2007 (UTC)

Probably would be the best option at this point. Seraphimblade ^{Talk to me} 02:38, 10 May 2007 (UTC)

If there was any mediation then the current vote is at 11 support (with some conditions in some votes relating to the first bullet) to 2 oppose. Fnagaton 10:52, 10 May 2007 (UTC)

It's not at all a "vote". They are comments about a proposal, which has changed since the first comment and which can and will be changed. Sarenne 11:06, 10 May 2007 (UTC)

You are wrong, it is a vote because there are votes for support and oppose. Fnagaton 11:17, 10 May 2007 (UTC)

There are "support" and "oppose", right. See WP:VOTE, it is not a vote (nor a straw poll). Sarenne 11:22, 10 May 2007 (UTC)

WP:VOTE does not support your argument and that means you're wrong, again. It is a vote. You also call it a vote here in this diff, so that means you're not being consistent again. Fnagaton 11:29, 10 May 2007 (UTC)

Oh, I forgot your argumentation "you're wrong, I'm right". Do you know what these " " are for ? I'm done "arguing" with you. Sarenne 11:38, 10 May 2007 (UTC)

You are wrong, the facts show it. Fnagaton 11:41, 10 May 2007 (UTC)

This is getting out of hand. I can deal with reasonable discussion on this (or any other) page. What I can't deal with is Sarenne's bullying campaign. This is highly disruptive to Wikipedia and it has to stop. You can't claim consensus when half the people disagree with you, and no one else is directly supporting your controversial edits. *** Crotalus *** 11:44, 10 May 2007 (UTC)

What is really disruptive (IMHO) is your reverts (and Fnagaton's, etc) without waiting for a new guideline, not my "initial" edits. When the guideline is changed (with a real consensus, not like what you did), I will do what the guideline recommends. Until then, I see no reason (IMHO avoiding edit wars is not a valid reason) why I shouldn't do what is recommended. This is my last answer here until I see an honest straw poll (that follows this guideline) about a proposal. All have been said.Sarenne 12:14, 10 May 2007 (UTC)

• Someone here has gotten some incorrect impressions of how guidelines work and are created. Specifically, according to policy, they are not created by voting on them, and according to more policy, consensus can change, so you shouldn't take a two-year-old consensus as indicative that there still is one today. Also, as WP:STRAW states, polls are never binding. >Radiant< 12:31, 10 May 2007 (UTC)

I support mediation.--agr 12:17, 10 May 2007 (UTC)

No matter how many times you call things a straw poll or link other policies that don't support your argument because it's not relevant here, it doesn't change the fact that we are discussing and working towards consensus. This has already been explained to you above when you tried to unilaterally block off the text. Fnagaton 12:35, 10 May 2007 (UTC)

• Actually, YOU may be, but Sarenne obviously is not, as xe has just stated that xe will keep revert warring over this until this page is voted upon. That is hardly a productive approach. >Radiant< 12:38, 10 May 2007 (UTC)

I think that should be grounds to permanently suspend Sarenne's account until the user promises not to revert war on this subject. Fnagaton 12:42, 10 May 2007 (UTC)

I don't know about suspending anyone's account, but I would strongly urge everyone involved to accept a voluntary moratorium on editing xB to xiB, or vice versa, until the mediation is settled. I certainly am willing to accept this. Seraphimblade ^{Talk to me} 13:01, 10 May 2007 (UTC)

• I do know about suspending people's accounts. There's presently several dozen edit wars on related articles. These must stop. Whatever the problem, edit warring is not the solution. >Radiant< 13:04, 10 May 2007 (UTC)

I will accept Seraphimblade's proposed voluntary moratorium on reverts related to this issue. I hope that Sarenne is willing to do the same. These reverts are not a productive use of time and I am tired of them. *** Crotalus *** 13:06, 10 May 2007 (UTC)

Anyone revert warring over this, in either direction, should be warned and blocked. It's disruptive and unproductive. — Omegatron 15:22, 10 May 2007 (UTC)

Users to participate in mediation

Before I open a mediation request on this issue, I need to get a list of which users are willing to participate. Some of the statements in the section above were ambiguous, and I don't want to include any users who would rather not be involved. If you wish to participate in this proposed mediation case, please sign below. After a day or so couple of days, I will ask for it to be opened at WP:RFM.

• *** Crotalus *** 12:50, 10 May 2007 (UTC)

• Yes. Also I'd prefer if Crotalus actually opened it. Also I think most active people here would want to be included and a day is too short. Fnagaton 12:53, 10 May 2007 (UTC)
• I will participate. Seraphimblade ^{Talk to me} 13:02, 10 May 2007 (UTC)
• I will too. --agr 14:55, 10 May 2007 (UTC)
• Me too. — Omegatron 15:19, 10 May 2007 (UTC)
• Certainly. -- mattb 15:57, 10 May 2007 (UTC)
• Same here --Marty Goldberg 16:01, 10 May 2007 (UTC)
• I will: -- SWTPC6800 00:44, 11 May 2007 (UTC)
• No problem --- Mahjongg 10:52, 11 May 2007 (UTC)
• Me too — jesup 12:00, 11 May 2007 (UTC)
• --SLi 00:56, 18 May 2007 (UTC)
• -- Shmget 15:10, 22 May 2007 (UTC)

What's going to happen during this mediation? Is it about prefixes in general or just the disruptive editors' behavior? Should we mention it on the talk pages of other people who have participate in this discussion in the past? — Omegatron 13:33, 11 May 2007 (UTC)

Hopefully, it would be geared toward finding a solution to the issue. Seraphimblade ^{Talk to me} 16:21, 11 May 2007 (UTC)

I was going to notify the other 6 or 8 people that have expressed an opinion about this topic on this talk page about the impending mediation, but when I started to make a list, there were actually... 85. See User:Omegatron/Binary prefix participants. A lot of people have opinions on this, and the handful taking part in the talk page as of today does not represent everyone. Should we notify them all? — Omegatron 14:55, 16 May 2007 (UTC)

If they are interested in the topic then they would almost certainly be aware of these discussions by now. Although I did a quick check of the list and the editors Crotalus horridus,Howard and TopaTopa seem to be missing whilst having expressed an interest in the binary prefixes topic. Fnagaton 00:12, 17 May 2007 (UTC)

Actually I find it quite possible that interested people are tired about tons of arguing in circles. --SLi 00:56, 18 May 2007 (UTC)

Mischaracterization by Crotalus

From [1]: "The above statement is clearly supported by the current discussion. If you disagree, show evidence that consensus to force binary prefixes exists NOW, not back in 2005."

There was never a consensus to "force" anything. The compromise reached in 2005 was that both styles are allowed, but, in specific instances, where accuracy is important, a change to standardized prefixes should be accepted. In addition, this page is a guideline, not a binding rule. In addition to that, WP:IAR always applies. — Omegatron 14:42, 10 May 2007 (UTC)

It's not "Mischaracterization", I suggest you retract your statement. This is because there is new consensus, up above in the vote. Fnagaton 14:46, 10 May 2007 (UTC)

What? Radiant's Crotalus' statement about the previous consensus is a mischaracterization. This has nothing to do with an imaginary "new consensus". — Omegatron 15:20, 10 May 2007 (UTC)

• I'm not sure what you're referring to, but I am rather obviously opposed to "forcing" anything, and well aware that guidelines aren't binding. >Radiant< 15:30, 10 May 2007 (UTC)

I was quoting your comment on the guideline page. You're claiming that the consensus from 2005 "forced" editors to use units in a way they don't like, but that is not the decision that was made. — Omegatron 15:53, 10 May 2007 (UTC)

Sarenne claimed the MoS gave him the right to force binary prefix changes in all articles. Fnagaton 16:05, 10 May 2007 (UTC)

Then take it up with Sarenne. — Omegatron 17:11, 10 May 2007 (UTC)

I and many others did, it is a matter of record that Sarenne refused to talk and instead carried on regardless, which has lead to the situation today. Fnagaton 17:23, 10 May 2007 (UTC)

Then go to mediation. That a single user is abusing a guideline does not invalidate the guideline. — Omegatron 17:41, 10 May 2007 (UTC)

About the current wording

At present, the page can be read to "authorize adding binary prefixes like [Sarenne] did" all over the wiki [2]. That's probably not the intent, and this may need addressing. >Radiant< 15:30, 10 May 2007 (UTC)

That's fine. Although using a consistent definition of SI prefixes throughout the project is desired by most editors, this should happen in an organic fashion, and not be forced against editors who want to use field-specific jargon in articles about that particular field. Of course this just means the debate will take place over and over again on every talk page of every computer-related article.  :-) — Omegatron 15:50, 10 May 2007 (UTC)

I've added language to try to close that loophole to make it more like what is intended by the above discussions/poll. Fnagaton 15:54, 10 May 2007 (UTC)

PS. I've checked the language used in the guideline and "editors should refrain from changing prefixes from one style to the other" and "follow the lead of the first major contributor to the article.". These alone would seem to not authorise the kind of changes Sarenne thinks he is allowed to do. Fnagaton 16:07, 10 May 2007 (UTC)

Oh yeah, please see the wording about years on this very page. "When either of two styles [AD/BC and CE/BCE] are acceptable it is inappropriate for a Wikipedia editor to change from one style to another unless there is some substantial reason for the change. For example, with respect to British spelling as opposed to American spelling it would be acceptable to change from American spelling to British spelling if the article concerned a British subject. Revert warring over optional styles is unacceptable". The same principle applies here. >Radiant< 16:13, 10 May 2007 (UTC)

Sarenne could say "some substantial reason for the change" applies because "it's a standards body that says so...". There needs to be something in the MoS which says "no" to plug that gap. Fnagaton 16:20, 10 May 2007 (UTC)

Actually (arguing with myself now!) no, the MoS now says there isn't consensus to use binary prefixes, so that means the "standards body said so" argument isn't enough to justify making all those binary prefixes changes. As the text stands it stops Sarenne making large number of binary prefix changes. Fnagaton 16:49, 10 May 2007 (UTC)

Some of the text you added is highly controversial, and I'd greatly appreciate it if you'd refrain from adding such text until there's some agreement to do so. Otherwise we'll be back to edit warring. I agree that everyone (Sarenne included) needs to stop making ANY changes related to this guideline for now, but adding disputed text to the guideline isn't the way to make that happen. Ask Sarenne to stop, don't modify the guideline just because of him. -- mattb 16:14, 10 May 2007 (UTC)

[AD/BC and CE/BCE] ... The same principle applies here.

Kind of. This is more than just a visual style issue, though. — Omegatron 17:15, 10 May 2007 (UTC)

Some folks here have a point of view that JEDEC standards on memory are irrelevant. They go so far as to point to a footnote in a JEDEC standard that has a copy of the IEC standard. This is used as proof JEDEC supports KiB and MiB.

A "relevant" standards organization is the IEEE. The IEEE Computer Society and all of their technical publications do not use the IEC binary prefixes. Here is sentence from an abstract in the May 2007 IEEE Transactions on Computers. "We describe and evaluate IMAP-CE, one of the latest IMAP processors, integrating 128 100 MHz 8 bit 4-way VLIW PEs, 128 2 KByte RAMs, and one 16 bit RISC control processor onto a single chip." [3]

This follows the "IEEE Computer Society Style Guide." Revised (October 2006) edition. [4]

The fact that IEEE uses the IEC binary prefixes is disputed and the IEEE reference should be removed from the style guideline. -- SWTPC6800 02:44, 13 May 2007 (UTC)

You realize that paper was contributed by researchers, not written by the IEEE itself. What's more, the "style guide" you link is nothing of the kind. It's a glossary which is internally inconsistent, so why you persist in citing it in this discussion is beyond me. The current guideline doesn't claim that all IEEE members use the IEC prefixes, only that the IEEE endorses them (which it does; it officially adopted the IEC standard after a trial usage). IEEE doesn't force all its members to conform to a certain style standard, but that doesn't change the fact that they endorse the style. -- mattb 03:01, 13 May 2007 (UTC)

Sorry Matt you are wrong. Read the IEEE Computer Society Style Guide Committee's mission statement. [5] They are very firm on their style guidelines.

"The IEEE Computer Society Style Guide Committee's mission is to clarify the editorial styles and standards that the Society's publications use. We maintain and periodically update a style guide to clarify those usages not adequately defined in accepted external sources. Our purpose is to promote coherence, consistency, and identity of style, making it easier for CS editors and our authors to produce quality submissions and publications that communicate clearly to all our readers."

It appears the IEC binary prefixes are so unpopular in the computer industry that the IEEE can't get their Computer Society to comply. -- SWTPC6800 03:33, 13 May 2007 (UTC)

That style guide defines "M" as the "SI prefix for million or mega". It makes no allowance for using "M" to mean 2²⁰, though it does make this allowance for "K". As I said, it is internally inconsistent and largely worthless. This isn't a matter of "getting the Computer Society to comply", it's most likely just a matter of whomever wrote that style guide either being unaware of or indifferent to the IEC prefixes. In any case, said style guide isn't at all consistent, doesn't represent good documentation practices at all, and isn't binding on IEEE-CS authors (of course, neither is the IEC prefix endorsement). Papers published in IEEE-CS journals typically reflect whatever binary prefix usage the authors please. -- mattb 03:50, 13 May 2007 (UTC)

Additional: I don't think anyone has claimed that the JEDEC endorses the IEC prefixes, but that's neither here nor there. The argument is that their utilizing common usage in their standards document (which they acknowledge as such) doesn't at all constitute an endorsement of prefixes. They standardize semiconductor memory, not unit prefixes. The claim that compliance with their standards extends to their glossary is an overinterpretation of technical documentation standards. The JEDEC never claims to standardize prefixes, and they point out in their definitions of terms that they are simply reflecting common usage. I'm not saying the JEDEC is irrelevant at all, but they haven't made any insistance upon or direct endorsement of any unit prefixes. They've just defined how they use them in their standards documents per typical technical documentation practice.

I've tried to contact the JEDEC on this matter so you can hear it from their mouths instead of mine, but thus far I've received no reply. Regardless, the claim that prefix usage (JEDEC standards) carries the same weight as a direct endorsement and recommendation of usage (ala IEC, IEEE, NIST, et al) is the main point I contest. It's nothing more than the common usage argument spun in an obscure fashion. -- mattb 03:17, 13 May 2007 (UTC)

Matt you're again making assumptions without evidence to support those assumptions. The JEDEC defines kilo/bega etc in the main body text of the standard and if you read the wording at the start of their standards documentation they demand conformance with all requirements (JEDEC JESD88B-3rd) in the standard, which means therefore they do endorse those terms. It is your assumption that they don't endorse those terms even though they are included in the main body text of the standard and that assumption does not make sense because you have supplied no supporting evidence for that assumption. It is much like your assumption about the IEEE Computer Society Style Guide and megabyte and gigabyte, just because you don't see a specific number definition you're making assumptions about what you think they mean, or you claim they are "largely worthless". Trying to claim things like that is not logical. Also throwing around terms like "common usage argument" in an attempt to somehow dismiss out of hand the arguments is also not valid. I could quite easily point out that consensus will often reflects common usage and then logically that means Wikipedia policy should eventually reflect the common usage consesus to not use binary prefixes. Fnagaton 09:57, 13 May 2007 (UTC)

And you seem to be extremely confused about technical documentation practice and the scope of standards. Either that or are just tickled pink that you've found a way to interpret a glossary such that it serves your purpose, I'm not sure which nor do I much care. The "common usage argument" classification isn't a dismissal as much as identification of what all these arguments are. The common usage argument is still valid, but there's no need to restate it a hundred times and insist upon it being some groundbreaking new evidence. -- mattb 16:42, 18 May 2007 (UTC)

I'm not even slightly confused about how to read a standard and I note that you are making snide remarks about me personally instead of tackling the real issue, this isn't the first time I've pulled you up on your attempts to use ad hominem. I am however pointing out you are making an assumption where you have not provided evidence to support that assumption. I gave you a chance to provide that proof, you didn't get it, deal weith it. But do not go repeating something you don't provide valid argument for, it doesn't make it true. The fact is there is new evidence. Fnagaton 22:16, 18 May 2007 (UTC)

You may be interested at the JEDEC representative's response on this matter below. -- mattb 23:51, 21 May 2007 (UTC)

A quick side note: From a quick look back at this page, it seems you (all of you, I'm not referring just to Fnagaton simply because his comment happens to be above mine) haven't made much progress, apart from more clearly agreeing to a mediation, since I stopped editing here 2.5 weeks ago (and there seemed to be consensus for mediation back then too). I still see the same arguments repeated. Of course agreeing on mediation is very good. --SLi 01:08, 18 May 2007 (UTC)

Except that there's been a change proposal, the text has been voted on, consensus was demonstrated and WP:MOSNUM has been changed to reflect this, all in the last 2.5 weeks. I think that is definitely more than "haven't made much progress". Fnagaton 08:47, 18 May 2007 (UTC)

SLi, while you were gone the topic was discussed on two other pages.

Wikipedia:Village pump (policy) - What constitutes "consensus" on the Manual of Style?[6]

Wikipedia:Administrators' noticeboard/IncidentArchive242 - Radiant's Bureaucracy Watch[7]

Here are a few quotes from people outside the debate.

• "Enforcing the manual of style" is not a justification for causing disruption.
• The guideline you're trying to "enforce" must not have a very strong consensus behind it, or you wouldn't be running into so much opposition.
• It seems to me that a small group of people want to change the fact that the world in general doesn't give a wet slap about the correct usage. Wikipedia is not the place to do that. End of story, I'd say. If the world outside is Wrong, then Wikipedia must be Wrong. And when the world gets it Right we can reflect that.
• Wikipedia is currently (to a first approximation, but a good one) consistent within each article, inconsistent across articles, and stable that way. Thus, Wikipedia reflects a real-world lack of consistency, and remains faithful to WP:NPOV by refraining from taking a side on the issue.

-- SWTPC6800 17:53, 18 May 2007 (UTC)

Merging computer unit articles

Since there are a number of people here with strong opinions and background on the issue, can I get your opinions on a related proposal to merge all the computer unit articles (kilobyte, megabyte, kibibyte, mebibyte) into one (or two) combined articles? See Talk:Kibibyte#Merge, for instance.

A while ago we decided to keep them separate and use a navigation template to tie them together, but I don't think this is really the best solution. — Omegatron 17:48, 10 May 2007 (UTC)

My suggestion would be to hold off on changes until the process here is settled. Likely we will end up with many links to those articles and what they say should be coordinated with what we come up with.--agr 18:31, 10 May 2007 (UTC)

The links to those articles aren't an issue. The merged article would still have the definitions of each unit, with an anchor tag to each, and the redirects would have anchor tags. So you could still link KiB, hovering over the link would still display "kibibyte", and clicking on the link would still go directly to the definition of kibibyte (except it would now be located at Data units#kibibyte or some such). It would just be a section or line in a larger article instead of an article of its own. — Omegatron 20:42, 10 May 2007 (UTC)

Sounds cumbersome to me. I thought it wasn't good practice to leave redirect links in articles. And someone who just wants to know what an unfamiliar term means gets stuck in the middle of a long article. In any case, I'm just saying any decision should be put off until the entire picture is clear.--agr 20:47, 10 May 2007 (UTC)

Going to run some experiments

I am going to create a couple of sandboxed versions of Commodore 128 in my userspace to test out various possibilities. I think that footnotes might be the best and least obtrusive way to go. *** Crotalus *** 22:19, 10 May 2007 (UTC)

Is this comment in the wrong section? — Omegatron 00:33, 11 May 2007 (UTC)

Yes, I should have opened a new section. Actually, after some thought, I've decided that Commodore SX-64 would be a better choice to test for 2 reasons: it's much shorter, and in addition to the standard usage, it also contains an ambiguous usage of "KB" (170 KB for the disk drive, which is actually a rough estimate on Commodore's part). *** Crotalus *** 00:48, 11 May 2007 (UTC)

Maybe floppy disk would be better? Some of the worst inconsistencies of terminology in that field. — Omegatron 01:32, 11 May 2007 (UTC)

You're right — that article is really a mess. But I think it's better to start with a simpler example before moving on to the most complex cases. *** Crotalus *** 01:46, 11 May 2007 (UTC)

Proposal

Take a look at this test article in my userspace. I believe this strikes a good balance, avoiding ambiguity while also using the same terms the original source material used. And, by using footnotes, it avoids clutter within the article text itself. Let me know what you think of it. *** Crotalus *** 03:18, 11 May 2007 (UTC)

The first thing I notice is that there should be a space before the "lb" and "kg" units in the first paragraph. :) Beyond that I haven't much to say. The notion of defining in every article what a unit means seems counterproductive and redundant to me. This might be a good solution for very obscure units, but kilobyte is used/misused so much that I don't believe this to be a reasonable route to take. Just my two cents, though, everyone else might love this idea. -- mattb 03:46, 11 May 2007 (UTC)

Well it's a better solution than doing nothing or changing everything to binary prefixes. ;) I do think it makes the article less cluttered to use footnotes. I do however think that the footnotes should be sprinkeled around the article text a little bit since they may be missed if it's just one. Fnagaton 10:41, 11 May 2007 (UTC)

I agree; probably footnoting both the first use in the infobox and the first use in the article itself (as we do currently with wikilinking for terms) would be a good idea. Obviously, the same footnote would be referenced in both cases, to keep the code size down. *** Crotalus *** 21:05, 11 May 2007 (UTC)

You missed the " for inch then, Matt!

Seriously, thank you for this experiment. I've pretty much kept out of this debate, although I do have a view on it. However, I have watched it from the sidelines, and I've been saddened to see the amount of name calling, accusations, assumptions of bad faith and so on from people on both sides (although not from everyone — some people have behaved in an exemplary fashion). But despite all that, I still like to think that a compromise is possible, and that we can reach a guideline which commands true consensus, as did eventually happen with the great year linking/unlinking debate. Some experiments are a worthy attempt towards that end.

As for the experiment itself, I think it may need some refinement. For one thing, the footnote defining "K" is in the infobox, which I tend to gloss over. Even if it was moved to within the main text, there is a problem that people often jump straight to a later section of the article and might miss the note. I'm wondering if it needs footnoting on every occurrence. (You can have a single footnote but link to it from several places).

Stephen Turner (Talk) 08:52, 11 May 2007 (UTC)

Adding a footnote to every occurrence of kB (or K) would quickly turn into a mess when it's often used. I agree that only using it in the infobox is a bad idea, once or twice in the main text should do it. Also the notation 2^10 may be technically correct, but I doubt everybody would immediately grab the significance. I would rather just see 1024 or maybe "2^10=1024".

Also, I notice that the reference about the disk capacity ignores the obvious fact that commodore was clearly using 1K=1024 bytes in it's literature, as 170x1024 = 174,080 which is very close to the real capacity. Mahjongg 10:50, 11 May 2007 (UTC)

That's true, but it's not exactly 170×2¹⁰ bytes, and since we do know the exact figure (and have a reliable source: the drive's user manual) that figure it should be listed. The purpose of this test is to attempt to meet the concerns of those who believe that legacy prefixes are too ambiguous. As to frequency of footnotes, I don't think we should footnote every occurrence, but it might be a good idea to do the first one in the article as well as the first one in the infobox. *** Crotalus *** 21:05, 11 May 2007 (UTC)

Sorry, I did not intend to say that the figure should not be listed, I intended my remark to show that Commodore was using 1K=1024 in their calculation, and given the context of this dispute it seemed a somewhat relevant observation. An observation that might be important enough to mention in the text. Mahjongg 01:14, 13 May 2007 (UTC)

Eh. — Omegatron 13:37, 11 May 2007 (UTC)

The above is a rather ambiguous comment... *** Crotalus *** 21:05, 11 May 2007 (UTC)

I personally like the parenthetical solution proposed earlier better, myself (though Crotalus' solution would be a good one when neither the binary prefixes nor the decimal ones properly express a value, that does happen in a few cases). As we get into newer systems, especially, it would be very tedious and problematic to have 13-digit numbers in exact values, and this would increase the potential for error in calculation. I do like the idea of listing the actual amount of memory available to the user, though. Seraphimblade ^{Talk to me} 01:41, 12 May 2007 (UTC)

Binary prefixes: A challenge

Over and over, people claim that the "common usage" of SI prefixes in a binary sense was consistent (it never was) until the evil hard drive manufacturers started using decimal prefixes to inflate their capacities.

• Crotalus: "In the 1990s, hard drive manufacturers deceptively began to use "megabyte" in the decimal sense — that is, 106 bytes — even though this clashed with common practice in the computing field."
• Centrx: "The hard drive manufacturers who apply the SI prefixes with their normal use are not doing so out of some concern over the accurate implementation of international standards, but in the interest of deceiving the customer for profit."
• Jesup: "It's true that HD manufacturers use powers-of-ten for advertising reasons"
• Rebroad: "As far as I can see, only hard drive and floppy drive manufacturers use a decimal million (10^6), primarily for marketing/misleading reasons ... I've worked in IT for 10 years, and I've never even heard of a Mebibyte"
• Cavebear: "im sure you can see where this is going. the ignorance of the general population and the desire for hard drive manufactures to write a higher number on their drives, does not change the correctness of the binary system being used to measure binary data."
• etc.

This sentiment even sneaks into our binary unit articles. I always suspected this was a bogus conspiracy theory, especially since it is always asserted strongly and without evidence, so I did some research. Now I'm pretty damn sure it's a bogus conspiracy theory. I don't think hard drives have ever been measured using the binary-meaning abbreviations. A short history of hard drive capacities, in the manufacturers' own words:

• 1954: "drum track ... 100 character positions" - IBM
• 1955: "drum ... 2048 words" - IBM
• 1956: "5-million-digit disk memory" - IBM ("byte" coined this year)
• 1957: "magnetic drums ... 60,000 characters each" - IBM
• 1963: "disk ... 100 million characters" - Honeywell
• 1963: "drum ... 2,621,441 characters" - Honeywell
• 1969: "disk ... 29.17 million bytes" - IBM
• 1976: "25 million" bits - DEC
• 1977: "nearly 15 million bytes" - HP

Up to that point they were not using 'SI' prefix either right... But in 1977 it was already common use that a KB = 1024 bytes... so you own 'research' show that the so called 'confusion' was introduced by hardware manufacturer, who started to use decimal meaning after the use of binary prefix has been established in the field. Shmget 03:03, 21 May 2007 (UTC)

• 1978: "67M byte formatted" (decimal) - DEC
• 1979: "67.4 MB" (decimal) - Fujitsu
• 1981: "10.66 Mb" (decimal, b = byte) - Quantum
• 1982: "5/10 megabytes" (decimal) - Seagate
• 1983: 9.57 "MB" (decimal) - Maxtor
• 1987: 21 "Megabytes" (decimal) - Seagate
• 1990: "1.216 gigabytes" (decimal) - DEC
• 1991: "1.53 GBytes" (decimal) - Micropolis
• 1994: "1,052 MB" (decimal) - Micropolis
• 2007: "One gigabyte, or Gbyte, equals one billion bytes when referring to hard drive capacity." - Every modern manufacturer trying to save their asses from frivolous lawsuits

Yet no modern manufacturer is using IEC notation, nor any of them 'explain' what a MB means when their refer to the cache... Point in case that the only possible confusion, which require a footnote in their specs, result of their odd decimal use when designating the capacity of their drives. Shmget 03:03, 21 May 2007 (UTC)

There's even an explicit "Maxtor adheres to the National Institute of Standards and Technology (NIST: www.nist.gov) definition of Megabyte and Gigabyte."

When was the transition from binary to decimal, again?

According yo your own research, there was no transition, just the beginning of mis-use of the units KB and MB circa 1978, at least a good decade after KB has been established as a unit meaning 2^10 bytes. Shmget 03:03, 21 May 2007 (UTC)

To save my poor eyeballs from further torment, I challenge someone to prove me wrong: Find a hard drive advertised using "KB", "kilobyte", "MB", "megabyte", or similar abbreviation, in which the unit has a binary meaning. If it was ever done, it was by no means common. — Omegatron 05:34, 14 May 2007 (UTC)

There is no need to do such thing to prove you wrong. The point being that the use of KB and MB as 2^10 and 2^20 predate the use 'decimal' MB by Drive manufacturer. They didn't 'change' their mind, they chose to use a misleading notation well after the conventional usage has been established. The marketing advantage of doing so has not escaped any reasonable observer of the market, and the current disregard of these manufacturer toward the almost 10 years old now, IEC notation, illustrate that their claimed attachment to the 'standard' is very circumstantial. Shmget 03:03, 21 May 2007 (UTC)

Please don't take my comment here as taking sides in this argument but, Omegatron, I have a question for you. How does the fact that manufacturers have never used the binary senses of prefixes to describe their hard drives imply that their motives not to do so were not to inflate the apparent size of their capacities? Just a question - I'm not about to wade into this. Jimp 06:13, 14 May 2007 (UTC)

It doesn't, Omegatron's "argument" is a strawman, as demonstrated by the language used such as "Over and over, people claim...". Fnagaton 10:15, 14 May 2007 (UTC)

In the settlement of Orin Safier vs. Western Digital the consumers got a limited time free download of a $30 software utility and the lawyers got $500,000. I like the footnote from Western Digital.

Apparently, Plaintiff believes that he could sue an egg company for fraud for labeling a carton of 12 eggs a "dozen," because some bakers would view a "dozen" as including 13 items.[9]

Western Digital must be using these IEC standards. :) -- SWTPC6800 02:54, 15 May 2007 (UTC)

*High five* :-) — Omegatron 23:50, 15 May 2007 (UTC)

Actually they do NOT, see http://www.westerndigital.com/en/products/Products.asp?DriveID=301 for example. "16 MB Cache - Bigger cache means faster performance. A massive 16 MB cache", and not 16MiB... Shmget 03:03, 21 May 2007 (UTC)

I believe the point was merely that a lot of people seem to assume that there's a marketing conspiracy going on, when really hard disks capacities have seemingly always been reported using powers-of-ten prefix notation. -- mattb 02:47, 21 May 2007 (UTC)

Please follow the link above to the IEC standard. :-) -- SWTPC6800 03:51, 21 May 2007 (UTC)

Yes, I got the joke. My comment wasn't directed at you so much as the classification of Omegatron's point as a strawman. -- mattb 23:37, 21 May 2007 (UTC)

Please move the dominant topic to its own page elsewhere

One single topic is dominating this page. Lively debate is good but I believe it is reducing visibility and debate of other topics. Please can somebody move the dominant topic to its own page elsewhere? Editore99 09:23, 14 May 2007 (UTC)

Good idea. Jimp 18:22, 14 May 2007 (UTC)

Unless there are any objections in the next few days I'll be moving the discussion to Wikipedia talk:Manual of Style (dates and numbers)/Binary prefixes. Jimp 03:15, 15 May 2007 (UTC)

Although I'm staying out of the discussion, I actually think it belongs here. Lots of subpages would make things harder to find later. Stephen Turner (Talk) 09:22, 15 May 2007 (UTC)

Plan A is to move it.

Plan B is to eliminate the multiple main headings. Make use of subheadings.

Plan C is to keep moving other topics to the bottom of the page.

Editore99 12:34, 15 May 2007 (UTC)

Sorry that the kilo/kibi debate has dominated this talk page for the last 4 months. It looks like the issue is nearing a solution. Give it another week or so before your vote it off the island. -- SWTPC6800 15:13, 15 May 2007 (UTC)

I'm not about to search through the archive but hasn't this debate been going on for years? Is it really ever going to end? I'm staying away too but I feel that a subpage (one not lots of them) would make things much easier to find: that whole debate would be kept together in one spot clearly linked to from here (as with the spelling debate at Talk:Aluminium) then it could be easily archived together (like /Archive 69) ... not to mention how much easier it would be to find stuff not part of that debate. Not so much a vote off the island but a vote onto its own brand-spanking-new island. I vote plan A. Jimp 15:38, 15 May 2007 (UTC)

As with other issues of this kind, like American/British spelling, the solution would be to allow both and not allow people to rampantly change them. The only issue there is that doing that defeats most of the purpose of using the new prefixes and exposes half their fundamental flaw. As with other issues of convention, the obvious solution is to use the most common spelling, but apparently that is refused as well. As with other parts of the styleguide, the styleguide never goes wholly opposite to the common practice on Wikipedia, but that is also refused. If others would follow the standard practice for all other issues of style, there would be no issue here. —Centrx→talk • 21:10, 16 May 2007 (UTC)

What constitutes "consensus" on the Manual of Style?

From Wikipedia:Village pump (policy) -- SWTPC6800 03:47, 19 May 2007 (UTC)

Way back in July 2005, a guideline was inserted into Wikipedia:Manual of Style (dates and numbers) on a 20-6-2 vote (the latter 2 arguing "No more stupid votes"). This was apparently done without consulting the editors of articles which would be affected by this change. As a result, discussions have flared up several times since then, with Manual of Style regulars generally favoring the status quo, and editors of the articles in question generally opposing the guideline. Recent discussion clearly shows that the guideline has no consensus at this point in time. I have stated on the talk page that I will remove it on the grounds of lack of consensus, and have been told in return that I can't do that, that a lack of consensus always defaults to the status quo. This seems wrong to me. A tiny handful of editors can make "consensus" on one corner of Wikipedia, and then enforce it everywhere and demand that others form a consensus against them before it stops? I do not believe the Manual of Style was ever intended for such purposes. Which interpretation of consensus policy is correct? Do MoS guidelines need consensus to remove, or is the lack of any consensus for keeping them enough to deprecate them? *** Crotalus *** 21:08, 9 May 2007 (UTC)

Regardles of anything else, the MoS is not supposed to be enforcable against consensus on a particular article as I understand it. It says "The guidelines here are just that: guidelines are not inflexible rules; one way is often as good as another, but if everyone does it the same way, Wikipedia will be easier to read, write and edit.". In general I do feel that a guideline version having obtained consensus (as this one apparently did back in 2005) a change should normally require consensus, but in an extreme case where there clearly is no consensus one way or another, a change to remove the guideline altogether might be warented, but i would hope for a broad-based discussion, not limited to MOS regulars, to establsih exactly what, if any, consensus now exists on this issue.DES ^(talk) 21:25, 9 May 2007 (UTC)

It's worth noting that the only reason this issue ever made it into the MOS in the first place (I was one of the persons who helped put it there) is precisely because leaving it to individual articles to decide wasn't working at all. The exact same debate with the same arguments would pop up quite often, each time requiring a lengthy fight to re-establish the same principles. In this case, the MOS guideline exists largely to centralize discussion on this matter. It's also a matter of consistency within Wikipedia. Is it a good thing for one article to use notation like "64K" to mean 2¹⁰ bytes and another to use "64 KB", another to use "64 kB", and yet another to use "64 KiB" to refer to the same quantity, all because of the collective preference of the authors of those respective pages? I don't think so, but you may disagree. Regardless, perhaps you can see why we wanted to centralize this to a place like the MOS.

I will also note that this sort of thing happens frequently as regards the MOS... Just look at the WP:MOSTM talk page to see how controversial enforcement of the "MOS consensus" on odd name capitalization/punctuation has been. That situation is a good analog to the binary prefixes issue; an editor will find an article like "the pillows" (capitalization intentional), fix it to use "The Pillows" per consensus on the MOSTM page (and correct English grammar), and a minor debate will arise on the MOSTM talk page. Since the regulars to that page endorse the current guideline, the article sticks to the letter of the guideline. The only real difference between the MOSTM issue and the binary prefixes MOSNUM issue is that the latter has finally gained critical mass due to a couple of editors who are willing to go through the weeks of debate required to give a change momentum. (upon further reflection, it's interesting that some of the arguments against using the binary prefixes could equally well be applied to the MOSTM capitalization/punctuation guidelines). -- mattb 21:46, 9 May 2007 (UTC)

This is an extreme case. The style states that if one contributor wants to use the optional style all of the other contributors must comply. There is one enthusiastic user that is making this optional style change in hundreds of articles. When the regular article editors complain they are directed to the WP:MOSNUM talk page. There they are told this style has the "consensus" and must be followed. The complainants out-number the "consensus" folks by a factor of 4 or more. -- SWTPC6800 01:11, 10 May 2007 (UTC)

Consensus is people agreeing to do stuff. If there are people in the field doing one thing, and there are some letters elsewhere saying another thing, then the consensus is with the people and the doing, not so much with the letters and the saying. :-) --Kim Bruning 02:37, 10 May 2007 (UTC)

The problem is that people aren't agreeing to do stuff. What has resulted is edit warring. One particular user claims justification under the Manual of Style to go around reverting changes, regardless of what the experienced editors on a particular article think. That guideline clearly lacks consensus. *** Crotalus *** 03:00, 10 May 2007 (UTC)

I think I said that. They can just revert the dude back, and if he's really fanatic, he'll just fall afoul of the edit-warring guidelines, (despite the Manual Of Style guideline suggesting elsewise ;) )

Either that, or just alter the text of the guideline. Go on, this is a wiki! be BOLD! :-) --Kim Bruning 03:04, 10 May 2007 (UTC)

Oh, if only such a rosy vision worked. This isn't an issue of article editors clashing with politicians who do no actual editing, but of editors clashing with other editors. As I stated above, this guideline came to be because the issue kept coming up in the course of editing several articles on different topics, not because a couple of us sat down one day and decided to try and make up a rule just for kicks and giggles.

I think the phrase "experienced editors on a particular article" is a rather loaded one. Nobody owns an article, and said "experienced editors" should be willing to talk through this conflict of opinion rather than getting miffed that some new person has changed "their" article to something they disagree with and insisting that consensus must be formed to follow the guidelines on the MOS. If the MOS has no teeth whatsoever, why the heck should we even bother with it? What if someone just doesn't like adding a space between a numerical quantity and a unit ("10m" as opposed to "10 m"). If we can just dispose of the MOS whenever we feel our way is better, what's to prevent me from finding some little nondescript articles that nobody takes an interest in and tailor them to exactly my preferred style of formatting? Sure the MOS should be applied flexibly, but I don't think that it's okay to just ignore it whenever an editor simply disagrees with it. We have fairly strict rules for the formatting of FAs, so at least some folks think that consistency is important. That's not to say the MOS can't and shouldn't change (and this guideline is in the long and arduous process of change), but the issue at hand is whether the current MOS text should be viewed as a consensus. I strongly believe it should. -- mattb 04:16, 10 May 2007 (UTC)

"This isn't an issue of article editors clashing with politicians who do no actual editing, but of editors clashing with other editors." I disagree. As far as I can tell, Sarenne has done no editing to computing articles (at least computing articles of the 8-bit era) except for changing styles. *** Crotalus *** 06:19, 10 May 2007 (UTC)

And I also disagree with your claim that "current MOS text should be viewed as a consensus". The alleged consensus was formed in 2005 on a 20-6 vote. That's 77%, which is consensus, but just barely. Since then, well over 14 people have complained that the guideline is stupid, makes no sense, violates WP:RS and WP:NEO, and so forth. Why do the opinions of those 20 people count more than the opinions of the numerous others who have commented afterwards, just because those people formalized it in a poll (which we're not supposed to do, anyway)? *** Crotalus *** 06:22, 10 May 2007 (UTC)

Without naming names, Sarenne isn't the only user involved in this with very few edits outside this prefix war. If you want to get all editcounty on this issue, the coin has two sides, but let's not go down that road. This is bigger than Sarenne since there are several editors who support the principle behind what he is doing.

Once more, if you're interested in numbers, you ought to note that most of the people involved in the current binary prefix debate were not involved in that vote you refer to. I'm not marginalizing any of the new contributors' opinions, but merely asking to continue centralized discussions rather than asking that the decision be farmed out to each article. -- mattb 06:35, 10 May 2007 (UTC)

User:Sarenne is just doing bot like edits on of hundreds of articles at a rate of 15 to 20 per hour. He has not demonstrated any expertise on these articles. The changes are just KB to KiB and such. You wonder why the "experienced editors on a particular article" get upset. This is not a conflict between editors, it is between editors and a gadfly doing WP:Point -- SWTPC6800 05:37, 10 May 2007 (UTC)

There's that ownership rearing its head again. Who says that he has to "demonstrate expertise" on an article to edit it? Who says that your expertise is greater than his? Who cares? There are tons of editors who make stylistic changes over a multitude of articles to conform with the MOS, you just disagree with this particular guideline and Sarenne is feeling the heat because he's applying it aggressively. It is unfortunate that this action has moved the cheese of several editors. However, Sarenne's actions have all been in good faith and are not in clear violation of any guideline, so I think it's rather unfair for you to accuse him of making a disruptive point. There are plenty of editors who agree with the principle behind the changes he's made (myself included), so let's not scapegoat the guy. -- mattb 05:38, 10 May 2007 (UTC)

I'm not claiming ownership of any articles. But article editing should work on a consensus-based process. If someone is making controversial edits and they keep getting reverted (especially if they are not a regular contributor to the article in question), they should discuss them on the article Talk page, rather than arguing that a policy has already been decided and they don't need to discuss the issue. *** Crotalus *** 06:19, 10 May 2007 (UTC)

I agree. There's almost nothing you can do on Wikipedia that you "don't need to discuss". Civility requires that if somebody asks you to stop and discuss, you indulge them respectfully, and take their opinion into account when deciding what "consensus" is. (Unless you're removing copyvios or reverting blatant vandalism or harassment, but that's clearly not going on here.) What's written in a guideline is just... something someone wrote down, and often fails to reflect consensus. There's no substitute for mindfulness. -GTBacchus^(talk) 06:29, 10 May 2007 (UTC)

So it is productive to re-hash the exact same arguments over and over? There's no merit in centralizing debate? -- mattb 06:35, 10 May 2007 (UTC)

Um, no. That's not at all what I meant. Perhaps I was unclear. It's profitable (I'd say necessary) to keep one's eyes and ears open, and to be alert to people's reactions. If you find that you're "enforcing" a consensus against frequent opposition, then it's probably time to revisit that consensus. That means precisely what you say: centralizing discussion, and helping to bring those who object to the guideline together with those who support it, at the guideline talk page, to figure out whether consensus has changed.

It seems like more work perhaps, but it's actually less work than dealing with the static generated by being stubborn. -GTBacchus^(talk) 06:40, 10 May 2007 (UTC)

I agree, and for this reason I can't condone all of Sarenne's methods. All I am arguing for is centralized debate and respect of prior consensus until it is changed, that is all. What I see essentially being proposed is license to ignore the result of a lengthy prior discussion simply because a user doesn't think it represents consensus. Is it for one person to decide that a consensus never existed on a matter? The implications of that disturb me. -- mattb 06:44, 10 May 2007 (UTC)

I agree about the centralized discussion. While there's a discussion that needs to be had, it's a bad idea to be making edits in either direction until that discussion happens. I'm not so much concerned with "respect for prior consensus" as with refraining from making controversial edits without discussing. That's disruptive, no matter who does it. I suggest a freeze on changing these prefixes until there's some agreement; no agreement means everyone should leave them alone, as we do with BC/BCE and "color"/"colour".

As for it being "for one person to decide that a consensus never existed", I don't get the impression we're dealing with just one person - am I wrong? If someone is disagreeing with the guideline, then whoever is "enforcing" the guideline should at least stop for long enough to point them to the appropriate talk page, and they'll either see that there really is broad agreement, or we'll all see that there isn't. I don't see what the hurry is to get the guideline enforced without pausing to talk about it. Communication is work, and it's worth it. -GTBacchus^(talk) 06:53, 10 May 2007 (UTC)

There are multiple users on both sides of the discussion (see WT:MOSNUM). As for the actual revert warring, the only person who has really been aggressive at forcing the disputed prefixes into articles is Sarenne, although this position does have a few other backers on the MOS talk page. These insertions have been variously removed by myself, Fnagaton, SWTPC6800, and Mahjongg. *** Crotalus *** 08:27, 10 May 2007 (UTC)

I'd just like to add that the whole issue of how enforcable the MOS should be is a very important one, but unfortunately it's one of those very difficult questions for Wikipedia that can lead to a perpetual debate and no clear answer (sadly). Oh well, there's the pitfall of direct democracy for you. (please resist the temptation to inform me of the NOT page... I've been around here way too long to subscribe to that rosy idyllic vision of what Wikipedia is and is not) -- mattb 05:55, 10 May 2007 (UTC)

I think that "agressively" is a good way to avoid applying style guidelines. It often generates more heat and disruption than it's worth. A better way to apply a guideline would be mindfully and with an openness to dialogue with other editors about why a particular style has consensus anyway. One always has to "apply" consensus keeping in mind that it might change out from under you; you have to keep tabs on how people react to a "consensus" that may be illusory, out-of-date, or who knows what. -GTBacchus^(talk) 06:02, 10 May 2007 (UTC)

It's my opinion that this particular alleged consensus is way out of date. The advocates of the status quo point to a poll taken way back in June 2005 that supported the guideline by a 20-6 vote. As soon as it actually began to be applied, complaints started pouring in. Whether or not there ever was a genuine consensus for this guideline across Wikipedia (I think there was not; most editors never read MoS pages), there clearly isn't now. *** Crotalus *** 06:14, 10 May 2007 (UTC)

(Before edit conflict): This is of course, totally true, and is why I don't personally go around changing a multitude of articles to conform with the MOS, only the ones I regularly work on. Still, I strongly feel that the MOS should be respected as consensus, otherwise there's very little point in keeping the useless mass of rhetoric around. Honestly, who cares what style we gently and cautiously suggest might possibly be a potentially good idea to use? In my own opinion, a style guide shouldn't be optional for something that we so boldly call an encyclopedia. Certainly the style guide should reflect the best practices, and therefore should always be subject to change, but accepting that we can just ignore it whenever we please renders it worthless. The question of whether the style guide needs ammending should be centralized, not a decision made on a per-page basis. These are merely my thoughts, take them for what they're worth or call me crazy for deeply believing in internal consistency.

(After edit conflict): The issues haven't changed since 2005, just the people exposed to them. This could very well mean that a new consensus needs to be formed, and that's totally fine. What I don't believe is fine is asking for license to ignore that consensus because we haven't yet clarified what the current one is. In any case, I fear that this huge binary prefixes debate is going to just bleed over into this page if we keep this up. When it comes right down to it, this is an issue of some editors asking to ignore a guideline because they don't think it represents consensus, and other editors asking for it to be respected because they do think it represents consensus. In between we get a ton of rhetoric and a lot of subtle and not-so-subtle finger waving and claims that the other party is "clearly wrong". Myself, I'm feeling rather ill at having written dozens of pages of text over a few little 'i's and would love to see this bloody thing converge to a conclusion, but I think people are happier debating than compromising. At this point I'm thinking that a binding decision is more useful than continuing to chase this elusive, mystical, and perhaps illusionary dream called "compromise". Too much time has been wasted on such minutia. -- mattb 06:35, 10 May 2007 (UTC)

Ok, so some editors want to ignore a guideline, and others want to enforce it. Until these people discuss and arrive at a conclusion they should all stop making these edits until the conversation happens. It's not about making a decision between "respecting previous consensus" versus ignoring it. It's about everybody stop editing until the conversation happens. Leave the articles as they are, whether that's right or wrong, and talk about it at MoS. It's that simple. Don't edit war, no matter how right you are. -GTBacchus^(talk) 13:18, 10 May 2007 (UTC)

The style he is changing is optional. His disruptive behavior depends on support of a handful of Manual of Style editors who claim the changes are mandatory. This is the only style guideline were a single user can force an optional style on all other editors. -- SWTPC6800 06:00, 10 May 2007 (UTC)

You know a style that is not "optional" ? As the MoS is currently written, all styles are optional.

These are not rigid laws: they are principles that many editors have found to work well in most circumstances, but which should be applied with flexibility. In this vein, editors should strive to have their articles follow these guidelines. While quality of writing may be more important than presentation and formatting, these elements also have their place in clear and unbiased delivery of information. One of the joys of wiki editing is that Wikipedia does not demand perfection. Wikipedia does not require writers to follow all or any of these rules, but their efforts will be more appreciated when they are guided by them.

The way I see the MoS is that they are recommandations. Writers are not required to follow these rules but if a single contributor change an article to fit the MoS, this change should (must) be accepted (except if there's a real strong reason not to do so). That's the only way Wikipedia will be consistent which is, I think, an important thing for an encyclopedia. Sarenne 10:49, 10 May 2007 (UTC)

You are the only person who feels strongly enough about this issue to go through articles changing it against the consensus on those articles. Stop it now. Your edits will continue to be reverted and all you will do is waste everyone else's time. *** Crotalus *** 11:02, 10 May 2007 (UTC)

There no "consensus" on all of those articles. You reverting all my edits is not a "consensus". Sarenne 11:16, 10 May 2007 (UTC)

I am not the only user who has reverted your edits. When you're in an edit war with four other people, time to step back and reflect whether you are out of touch with current consensus. *** Crotalus *** 11:19, 10 May 2007 (UTC)

I won't "step back" only because 4, 5 or 10 contributors don't want to wait for a new consensus and are trying to revert all my edits to make a point, against the current guideline. If you think that there's a new consensus about a new guideline then talk about it there : WT:MOSNUM. Sarenne 11:32, 10 May 2007 (UTC)

We have been discussing it, and keep repeatedly running into brick walls of bureaucracy. What you don't understand is that the MoS guideline only has any force if it represents consensus. As of this time, it doesn't. What you are doing is simply disruptive edit warring. *** Crotalus *** 11:38, 10 May 2007 (UTC)

Sarenne, hi. I agree that you shouldn't "step back" because 5 contributors don't want to wait for a new consensus. I'd step back because it's the right thing to do. Once a dispute arises, the only correct action is to stop editing and talk to the people. If the consensus really is as you say, then it won't take long to convince people of that, and if not, then we should find that out quickly. Reverting without discussion is never right; because it's never civil. Think about that. -GTBacchus^(talk) 16:57, 10 May 2007 (UTC)

I don't have to think about that, I know that. You (and apparently all of you) think it is the right thing to do, I think the right thing to do is to prevent users from reverting again and again edits that follows a guideline until the guideline is changed. I tried the discussion, again and again (and again (and again))... If you stop editing when there is a dispute about "binary prefixes" then you'll never edit and the guideline is useless. That's why the "encyclopedia" is inconsistent : guidelines are worthless, they have no teeth. Sarenne 17:15, 10 May 2007 (UTC)

Being consistent about binary prefixes is not more important than refraining from edit warring. If you want to "prevent users from reverting again and again", then you need to stop reverting; otherwise you're edit warring to stop an edit war, which is always wrong. It's like bombing for peace, or screwing for virginity.

The guideline you're trying to "enforce" must not have a very strong consensus behind it, or you wouldn't be running into so much opposition. Your job, and the job of those opposing you, is to stop editing, talk on the guideline talk page, and determine what the consensus is now. It doesn't matter what it was two years ago; it matters what it is now. Since it's now in dispute, your editing to "enforce" it is inappropriate.

It's this simple: Once you know there's a conflict, stop and talk. We're not in a hurry, but we are under an obligation to be excellent to each other always. That means listening, and trying to respond to current consensus as you detect it. Consensus is not detected by reading a guideline, but by listening to editors. If there's no consensus, then editing binary prefixes in either direction is inappropriate, just like we don't edit "BC" vs "BCE" or "color" vs "colour". -GTBacchus^(talk) 19:30, 10 May 2007 (UTC)

I totaly disagree with all that you've just said ;). Being consistent about binary prefixes (or an other style) is more important than refraining from edit warring. If you stop whenever there's a conflict about a guideline, you make Wikipedia inconsistent (and sometimes unstable), that's exactly what happened with "BC" vs "BCE" or "color" vs "colour", except that with the current "guideline" about binary prefixes we'll have both ambiguity and inconsistency (and maybe instability). It's weak rules that cause edit wars. What you've said may be the right thing to do for a content dispute (about the meaning) but not for the style, which should be centralized and always binding. For the style, we need strong rules, not "guidelines" that you can suspend whenever there's a dispute, which makes them totally worthless. You are too afraid of binding rules. Of course, these rules can change but until they are changed (with a consensus), a user who follows the guideline should never be blocked, even if he's edit warring. I'm done listening and discussing about this guideline. Four month are enough. All have been said. I know that no one will agree with me but that's what I think :) Sarenne 21:25, 10 May 2007 (UTC)

(de-indenting) Better than simply disagreeing with you, I can explain to you why you're wrong. :) The "BC" vs "BCE" issue is one with which I'm pretty familiar, having worked on it in the past. Wikipedia is currently (to a first approximation, but a good one) consistent within each article, inconsistent across articles, and stable that way. Thus, Wikipedia reflects a real-world lack of consistency, and remains faithful to WP:NPOV by refraining from taking a side on the issue. We just leave the date formats alone once they're consistent within an article, unless there's a compelling reason to use one or another in that particular article. It's stable because people know not to mess with it, and that works.

As for consistency being more important than not edit warring, that makes no sense - allow me to explain why. We work towards consensus because this is a wiki. It's inherent in the software that we can't just go around enforcing our ideas against significant disagreement. Anyone can edit, and if you go against a lot of people, they'll edit it back anyway. Since you can't control them, you have to discuss. It's not a rule so much as a law of nature: if you don't swim, you're gonna sink. The name of the game here is consensus, and we have no choice about that.

It's not about "fear" of binding rules, it's about how we work together as human beings and get something done. If you think Wikipedia needs more binding rules, then I think you can find other online encyclopedias that work that way. They're not nearly as successful as Wikipedia, which is why I think the "No binding rules" philosophy is actually pretty effective. Edit wars are caused by edit warriors, and they're always wrong.

Edit warring is bad because it makes article histories and "recent changes" less useful, it distracts editors from getting productive work done, and it encourages others to edit war, over style, content, and everything else, leading to a Wikipedia that's bogged down in back-and-forth, "is not!"/"is too!" arguments. The only civilized solution, the only solution that works on a wiki, is for everyone to work for consensus. Yes, that means stopping and talking. Yes, that makes things take longer. No, we're not in a hurry. No, it doens't make guidelines "worthless"; it makes them more responsive to the community and better indicators of consensus. Yes, we all learn more and respect each other more in the process. -GTBacchus^(talk) 00:40, 12 May 2007 (UTC)

You're IMO describing an utopia. We have binding rules and we don't suspend them if in a talk page it seems there's a lack of consensus about it. After an edit we don't "stop and talk" every time we disagree (maybe you think we should).

Policies and guidelines should reflect the consensus, then we should always assume than they reflect consensus until they are actually changed. We should "stop and talk" only if it doesn't concern a guideline or a policy. We don't "stop and talk" when dealing with vandalism. We shouldn't "stop and talk" when dealing with style, we should talk... and stop only if there's a new strong consensus or if there's no guideline. IMHO that's a pragmatic application of the "consensus spirit". Sarenne 10:46, 12 May 2007 (UTC)

Hmm. You'd be surprised how often things work just the way I've described. You'd be surprised just how non-binding our guidelines are. Stick around, and keep your eyes and ears open, and you'll find out a lot. Policies and guidelines should reflect the consensus, true, but when there's some indication that consensus may be changing (as there is in this case), how are you going to know the consensus has changed until that conversation happens? You'd be surprised how much time it doesn't waste.

Your opinion about pragmatic application of the consensus spirit is not borne out by experience, in my view. Of course we don't stop and talk when dealing with vandalism; that's entirely different. However, when good-faith contributors disagree, that's what talk pages are for. We are not in a hurry. It is certainly less than respectful to continue in an action that you know a group of people disagree with, without engaging them in some kind of discourse.

The truth of the matter is that taking the time to be certain of continuing consensus makes one's edits much more sticky, obviates the perceived "need" for edit warring, contributes to a more civil and collegial environment, and most to the point, works. If you don't believe me, try it. I've "won" plenty of disputes, always by refusing to edit war, and pursuing other strategies instead. They're very effective, those other strategies. -GTBacchus^(talk) 18:29, 12 May 2007 (UTC)

An afterthought... the edit-warring, enforcement strategy that you want to pursue... it leads here, to long conversations at the Village pump. Not very pragmatic after all, is it? -GTBacchus^(talk) 18:32, 12 May 2007 (UTC)

I've been bold and changed WP:MOSNUM#Binary prefixes to reflect what I believe is the current (lack of) consensus. Basically, I copied some of the wording from the "National varieties of English" section of WP:MOS, and suggested that articles should stay with established usage (similar to "Stay with established spelling") and follow whatever prefixes were used by the first contributor (similar to "Follow the dialect of the first contributor"). There is clearly no consensus to mandate binary prefixes, regardless of the outcome of a 2005 vote. Likewise, I doubt there is consensus to get rid of the neologisms entirely. Therefore, all that can be done is to make a guideline that attempts to stem further edit wars of the sort that have been fought over the past several days. *** Crotalus *** 11:18, 10 May 2007 (UTC)

And I see that Sarenne already reverted it. Very well, I won't edit war on the MOS page; but I maintain that there is no consensus whatsoever for the current alleged guideline, and that Sarenne's repeated edit wars in the face of opposition from numerous other editors is highly disruptive to the encyclopedia. *** Crotalus *** 11:30, 10 May 2007 (UTC)

You don't change guidelines just because someone is using them as justification for an edit war. That's what WP:3RR is for. Take it up with the person being disruptive. — Omegatron 21:52, 10 May 2007 (UTC)

The whole point of a Manual of Style is to suggest a uniform style for the entire project. It doesn't "enforce" anything.

Consensus can always change, and "consensus to change a guideline" or "consensus to demote a guideline" is a bogus idea. Policies and guidelines don't become "stuck" after a small number of people have agreed on them. As soon as editors stop agreeing on something, it is no longer binding. — Omegatron 14:51, 10 May 2007 (UTC)

Absolutely. Being written down in a guideline doesn't make something set in stone. The guideline reflects what we're thinking, and often lags behind. It certainly holds no authority unless it's supported by continuing consensus. -GTBacchus^(talk) 19:34, 10 May 2007 (UTC)

Exactly. This "sticky consensus" concept is absurd, and getting out of hand. When there's no longer agreement for something, there's no longer agreement for something. Simple as that. — Omegatron 21:52, 10 May 2007 (UTC)

This is not absurd at all. It's always easier to remove something with an apparent consensus than to build something with a consensus. For example, the readers or editors who don't "like" binary prefixes will naturally go to WT:MOSNUM to discuss the guideline whereas the readers who "like" them have no reason to go there. That's why we should always have a consensus to suspend/remove/change a guideline but not to keep it. Sarenne 22:07, 10 May 2007 (UTC)

Message to GTBacchus, there are alot more then just five of us as a matter of fact we (user who think MB is more exceptable then MiB) greatly outnumber the ones like Sarenne.--  Planetary Chaos  Talk to me  17:50, 10 May 2007 (UTC)

You have no idea how many people are one side or another (or another or another). You only know about the few vocal ones on the MOSNUM talk page. How many of the people who participated in the original 2005 discussion are contributing to the current discussion? Why do you think that is?

Even if you did know exact numbers, Wikipedia is not a democracy, and a +1 majority doesn't completely invalidate the minority's opinion. Everyone's positions must be taken in to account. — Omegatron 21:52, 10 May 2007 (UTC)

I wasn't trying to claim that there are only five or only fifty of you. My only point was, even if it's only five people opposing, you stop and talk to them, because Wikipedia fails when we start deciding that discussion is somehow unnecessary. -GTBacchus^(talk) 19:34, 10 May 2007 (UTC)

• Why does this debate always seem to swiftly degenerate into one of “Let’s keep the old consensus (because it supports my way)” versus “Let’s change it (because it supports their way, but not my way)”? There are “third ways” – that have been proposed to deal with this very issue – which accommodate both ways without doing harm to either way. Given that, just what is it – besides the hope of still making a point in one’s favor – that keeps the bitter vile flowing? Askari Mark (Talk) 04:27, 19 May 2007 (UTC)

Consensus ?

Wikipedia seems to be used to promote a minority POV trying to enforce one notation of the IEC. Yet 9 years after the 'adoption' of this IEC directive, The usage of these new symbol is still confidential. The only notable exception seems to be on Wikipedia, where a motivated group have apparently decited to teach the rest the industry how wrong they are.

Amusingly enough, the industry players that should have welcome the IEC directive, namely the Hard Drive manufacturers, still do not use these units in their spec sheet (they still refer to cache memory in MB, and they still are compelled to explain their peculiar meaning of MB/GB for when referring to hard drive capacity with a footnote explaining that in this context a GB is 1 billion of byte (there are no such footnote to explain that a 8MB cache is 8388608 bytes. Some open source programs have offered their users the possibility to use the IEC notation, yet of course, none of them have change their default behavior. Every major Industry players, from Intel to AMD and Dell still used MB in their historical and conventional meaning. see http://www.hitachigst.com/tech/techlib.nsf/techdocs/413CA7F7E4F76D3986256D270072AF9E/$file/15K147_Final-update.pdf , http://www.seagate.com/staticfiles/support/disc/manuals/desktop/Barracuda%207200.10/100402371f.pdf ,

Dell in his marketing brochures use MB and GB for memory without further comment, but their is a footnote next to the size of disk, that refer to a page that explain; "For hard drive, GB means 1 billions bytes and TB equals 1 trillion bytes: actual capacity varies....." see http://configure.us.dell.com/dellstore/config.aspx?c=us&cs=19&l=en&oc=DXCWNE3&s=dhs , see 6th bullet down the page.

It is notable that Dell do not bother to explain anywhere that MB, in other context that the one annotated, means 1024KB, despite spending two full page to a bunch of explanation notes, Dell, one of the major PC retailer, whose marketing mail-in material is definitely target to the non-professional public, assume that everyone understand what a MB really is. http://configure.us.dell.com/dellstore/config.aspx?c=us&cs=19&l=en&oc=DXCWNE3&s=dhs , see 5th bullet down the page, that explain that to fully exploit more than 4GB of memory you need a 64Bits OS... without ever botehring to explain what GB is... adn of course without botthering with GiB. Note that I picked Dell, because I happened to have a printed copy handy, but that observation is true of any major PC retailers.

I have noticed, browsing some edit, that some will take the stance that Seagate, Hitachi, Dell, Intel, AMD, IBM, Microsoft, etc... are all WRONG, and that the industry de facto standard should be ignored....

The same picture can be drawn from the software industry. The use of IEC notation is the rare exception, and when it exist it is a GUI option, and to the best of my knowledge never the default option. For backward compatibility reason, the many syntax that accept K,M,G as suffix to number cannot poosibly cahnge. They may accept both K and KiB, but will not cahnge the meaning of K as 1024 or they would create a huge backward compatibility breakage. That mean that most object defintion language of most Database software, for example, will not change the established usage, even if every article of Wikipedia claim that they are 'wrong'.

Norms are good, and good norms get adopted quickly, especially when they solve a problem. The IEC norm can have some legitimate usage, but Wikipedia, or an encyclopedia in general, has no vocation in pushing what some consiter the 'correct way(tm)'. That there should be article explaining the IEC norm and this history behind it is one thing... parsing Wikipedia to systematically chage all KB,MB reference, that is a crusade not an encyclopedic work. The IEC recommendation has been out there for almost a decade now... Not all normative attempts are successful, and it should not be the place of Wikipedia to be a tribune to promote it's adoption by the industry and the rest of the world.

As far a the 'manual of style' goes, I find it disturbing that the consensual usage of the profession be arbitrarily over-ruled. It would be like suppressing all references to carats in every articles about Diamonds.

Bytes and therefore KB,MB,TB are NOT unit of the SI. The meaning of these units IS an industry de facto standard and is pervasive enough that they do not need explanation in non-specialist marketing document, just like carat is relating to diamond. This relate to WP:NOT#DICTIONARY: "Wikipedia is not in the business of saying how words, idioms, etc. should be used.". Wikipedia should be the reflect of the actual usage, not what some think it should be.

For all these reasons, I think that the Manual of Style should not express the view that the conventional KB, MB usage is 'wrong' not should it mandate the use of IEC notation, and especially not promote or support a campaign of massive edit in favor of IEC notation. Shmget 20:04, 20 May 2007 (UTC)

To quote Guy "It seems to me that a small group of people want to change the fact that the world in general doesn't give a wet slap about the correct usage. Wikipedia is not the place to do that. End of story, I'd say. If the world outside is Wrong, then Wikipedia must be Wrong. And when the world gets it Right we can reflect that." [10] I agree. -- SWTPC6800 04:50, 21 May 2007 (UTC)

I doubt the world in general gives a wet slap about correct usage of a space between unit and value, but we seem to. This is a non-argument, there's a balance to be struck between common usage and unambiguous encyclopedic writing. Anyone trying to simplify the issue to black and white, (I'm) right and (you're) wrong, is either silly or far too empassioned IMHO. -- mattb 23:42, 21 May 2007 (UTC)

If a typical encyclopedia reader encounters "100 mm" they generally will not have to look up the significance of the space between the number and units. If they read "8 mebibyte", they most likely will need an explanation. -- SWTPC6800 05:06, 22 May 2007 (UTC)

Yes, and unfortunately they often also need an explanation when they see "megabyte" (whether they realize they do or not). If this weren't the case we wouldn't even be here. -- mattb 05:09, 22 May 2007 (UTC)

Well, I can only speak for myself, but the reason I'm here is NOT because I'm confused about the meaning of MB, nor because I believe that there is any risk that a computer owner may not know that 4x512MB = 2GB. (begin rant) I ended-up here in the quest to understand how a marketing scam, that started in the late 70's early 80's ended-up being described as the saint victim, and how articles end-up describing the universal usage of memory unit as power of two as 'caprice of computer nerds' came to be. I actually came here from the French Wikipedia ( I straddle the english and the french one), where Sarenne is running the Standard Police, editing pages faster than a bot in a frenzy of historical revisionism, disfiguring articles on machines that stopped being produced at a time where a Hard drive was an exotic and insanely luxurious device for a 'home computer'. And no, universal usage of memory unit is not an exaggeration, even the biggest offender - the hard drive manufacturers - still, to this day, refer to the amount of cache memory of their drives in MB, without any footnotes. and it IS computer MB. That the BIPM is now cornered and has no other choice than to stick with M = 1000K, is one thing, but claiming that this had precedence, and that 'Harddrive manufacturer have 'consistently' been using it, way before memory makers and the rest of the profession, is just mind boggling. It's not like it was in a distant past, I, and many other still around, have lived through it, have witnessed the events unfold first hand. I still own a 8bit machine with 64K of memory and a 140KB (yes KB as in 1024B) floppy drive, I still have all the manuals, magazines, books of the time. And at no time, ever, did anybody ever mentioned being confused about KB and MB. Actually hard drive vendor at the time where tap dancing around the MB notation, using all kind of wording in Ads (like 5 Megs 5 M. bytes, 5 M Ø) . Sure at the time the target audience would have called bullshit immediately had they been using fake MB, later in the 80's with the popularization of computers and huge growth of the costumer base, they finally were able to play that dirty trick... But now, the story have become... 'the poor hard drive maker just did what was good and right from the beginning, but these nasty computer nerds have seeded confusion corrupting the purity of the I.S.'... As they say in Grey's Anatomy: 'Seriously!'. Seriously, I'm sure that there is legitimate use of KiB in some articles, and even maybe in 20, 30 years when all that is left is the Java Generation of programmer - whom the only thing they know about memory is that they don't have enough of it - maybe then KiB will be the natural thing to do, but in the mean time I am torn between cry and hysterical laughter when I read a page that claim that an Apple ][ had 48 kibibytes of memory!!!! why not rename the Indy 500 into the Indy 500 Statute miles, we would not want the public to risk confusion with the I.S sanctionned Mile (symbol M) which is a nautical mile (end rant) -

Now, practically, in the spirit of conpromize, here are my personal limits:

* A footnote, or why not a template, to precise the usage/meaning of KB of the page... well, fine. I'll get over it.

* A double notation like Cache memory 4 MB (MiB), again fine. a bit heavy on the eyes, but heck , I'll get over it.

* Replacing KB with KiB, especially in article where there is no chance in hell that anybody interested in the article be confused, Nah!

* Using 'kibibytes' anywhere but in some specialized norm-fanatic hair splitting heaven, Hell No(1). use the wikified abbreaviation KB (KiB).

(1) at that point I'd write a firefox extention to automatically filter them out of display. Actually that make me think: wouldn't be possible to make that a User preference, like the Date format ?