Wikipedia talk:Manual of Style (dates and numbers)/Archive B17

binary SI prefixes vs decimal SI prefixes

 * NOTE FOR ARCHIVING: When this thread is no longer active and has timed-out, please create a B17 archive for this thread. Please also list it under the “Binary prefixes” classification of the Archives.

The SI should be recognized as one of the most widely adopted standards in business and science, and as such, should be followed when it comes to the meaning of SI prefixes, rather than following the use of kilo, mega, giga as powers of 2, which is confined to some parts of the computer industry. Nobody outside of engineers in the computer industry has any concern for calculating in powers of 1024. It's not only useless, it takes more time to make calculations, and can lead to confusion or errors. There are about 3 countries worldwide that have not yet adopted officially the SI: Burma (Myanmar), Liberia, and the United States (International System of Units).

NPOV is there to protect against abuses, not to make sensible changes impossible. In this case, it's a really good choice for coherence, and a greater consistency of units. I gain exactly the same as anyone else: better units and coherence. As a consequence, I think we should entirely remove the column "Binary Usage" of the templates "Quantities_of_bits" and "Quantities_of_bytes". At the very least, mark that column as "confusing" and/or "deprecated". Compvis (talk) 05:12, 25 April 2010 (UTC)


 * It is not true that "nobody outside of engineers in the computer industry has any concern for calculating in powers of 1024." Many people use computers for all kinds of purposes, and it isn't unheard of for non-engineers to need to know exactly how many bits or bytes are in a file, or will fit in a device. Since the computer industry has adopted the powers of 1024 meaning of prefixes, all users of computers are forced to understand these meanings. An alternative proposal exists, using prefixes such as kibi-, but the proposal has largely been ignored. It is not for Wikipedia to adopt a proposal that the general public has decided to ignore. Jc3s5h (talk) 06:19, 25 April 2010 (UTC)


 * "it isn't unheard of for non-engineers to need to know exactly how many bits or bytes are in a file" =>Indeed... In that case, if they want some precision, they wish to have this number in decimal to make calculations easier! If they are geek enough to want to have the exact number, then even if you try to stop them, they will still figure it out. Base 1024 doesn't have any advantage, whereas base 10 is a worldwide standard. Still, you haven't adressed any of my arguments, SI prefixes ought to be powers of 10 (for convenience, and to follow the most widely used standards related to the subject), and it's an historical error that it's sometimes hasn't. When will people say "Yes, we screwed up, kilobytes ought not to have 1024 bytes, just like kilometers ought not to have 1024 meters, and the same for kilograms, kilonewtons, kilopascal, etc..."?.


 * For others arguments, please see "SI is one of the most..." Compvis (talk) 03:30, 27 April 2010 (UTC)


 * This has been debated to death time and time again, with several people getting blocked over this due to the all-out warfare mentality of several participants. See the 16 MoS archives dedicates to these discussions (B1–B16, linked in the archive box found in the top right corner of the page), as well as countless RfCs on it. When the computer industry decides to stop using K, M, G, etc... to mean powers of 1024, so will Wikipedia. So unless you can show that this practice is significantly less popular today than last time, you'll won't get much enthusiasm to re-open this issue as most will consider this to be a WP:DEADHORSE. Headbomb {talk / contribs / physics / books} 04:14, 27 April 2010 (UTC)


 * Did you ever consider that wikipedia gets the first place for a lot of searches? Of (kilo, mega, giga, tera) X (bit, byte), the only case where wikipedia is not the first result on google is gigabit, in which case it is fourth. Do you expect this madness to stop? Don't you think the industry justifies their stance on the units by saying "look at all these people that use it that way (including a lot of people on wikipedia)"? Apple has seen the light on this matter. With all due respect, I consider this to be a WP:LACK_OF_UNDERSTANDING_OF_THE_WORTH_OF_CONSISTENCY_IN_UNITS, and a WP:DESIRE_TO_LET_THINGS_GO_BACK_TO_MIDDLE_AGES_WHERE_EVERY_UNIT_WAS_KINGDOM_DEPENDENT . Compvis (talk) 13:12, 27 April 2010 (UTC)
 * Wikipedia is not a soapbox upon which to promote your view of how things OUGHT to be. Wikipedia merely describes how things currently are; and the fact is that (at least to my knowledge) the binary use of the SI prefixes with respect to quantities of data is still (unfortunately) current widespread practice and the IEC -bi- units are virtually unused (sadly). Also, this is a style guideline, so discussing the template is off-topic IMO (I cannot even begin to follow the logic of the editor who suggested that you bring your discussion here). Now, if you wanted to suggest something like following the IEC's usage suggestions when reporting data quantities in Wikipedia articles, that would be relevant (and I for one would support it). --Cyber cobra (talk) 14:35, 27 April 2010 (UTC)
 * Interested parties might want to read this Dondervogel 2 (talk) 18:23, 27 April 2010 (UTC)
 * Yes we're all aware of things. I'm personally pro-IEC prefix. However I also recognize that's a minority use. Headbomb {talk / contribs / physics / books} 23:16, 27 April 2010 (UTC)
 * Ok I read the soapbox article. The question that people should ask is the following: If "content hosted in Wikipedia is not for: Advocacy, propaganda", then why is Wikipedia, by trying to describe "things as they are" (and being ranked so well), "promoting/advocating" the status quo? The real problem is that nothing is neutral, or even close to being so. Wikipedia promotes status quo. Thank you Wikipedia, we really need more confusion in the units, as if it hadn't been hard enough to get the SI adopted. I like having three likely meanings for mega! Can i haz more?


 * Regarding the link to Thunderbird2's talk page that user was just about to get blocked for disruptive editing and pushing a minority point of view against consensus. The consensus is that Wikipedia is not a soapbox, it therefore represents that world as it is. In this specific topic the world has rejected using the SI and IEC prefixes as those organisations advocate.Glider87 (talk) 09:02, 28 April 2010 (UTC)
 * "Rejection" is a tad strong. More like "failed to adopt". --Cyber cobra (talk) 09:45, 28 April 2010 (UTC)
 * Yes, WP should describe the world as is, but this should be in a way to avoid misunderstandings by the non-initiated reader. Therefore disambiguating ambiguous units would be valuable. And since an international standard is available, why not use it for that purpose? &minus;Woodstone (talk) 09:52, 28 April 2010 (UTC)
 * SI/IEC isn't the disambiguation standard used in the real world. This means to use SI/IEC here would be pushing a point of view and soapish. This is exactly why they shouldn't be used here on Wikipedia. In the real world if disambiguation is used then the number of bytes is written. Glider87 (talk) 10:05, 28 April 2010 (UTC)


 * To put it simply, WP:JUSTDONTLIKEIT or WP:ILIKEIT for advocating SI/IEC is not an argument suitable to use in talk page discussions. In the world SI/IEC use is strictly in the minority therefore they are not to be used here except in the most rare and specific cases when they are specifically the topic of articles. Also in the world SI/IEC use for disambiguation is also strictly in the minority therefore that methpod is not to be used here either. Wikipedia uses the methods demonstrated by the real world and this means using the prefixes KB and MB and others in the binary sense or decimal sense. This also means that if disambiguation is needed then the real world method of writing the precise number is to be used on Wikipedia. Glider87 (talk) 10:36, 28 April 2010 (UTC)


 * Compvis, once a comment has been left up and responded to it is customary to use the strike feature if you want to retract some text instead of removing it completely . This preserves the talk page flow and context for the archives. I have put back the text using the strike tags.Glider87 (talk) 22:24, 28 April 2010 (UTC)

Use of precise numerical value provides a very good disambiguation. But by insisting on that as the only acceptable disambiguation, Wikipedia has shot itself in the foot. (I suspect that, by deprecating IEC prefixes, the present wording is interpreted as a licence not to disambiguate). To put my point a different way, to insist on an exact number of bytes poses a high effort threshold, and most editors probably cannot be bothered. It is (much) easier to add the missing “i” in “GB” where needed than to type a complicated footnote explaining that “GB” sometimes means one thing and sometimes another, in every single article. The task of converting to precise values, if that is really necessary, can then be carried out by a bot. Dondervogel 2 (talk) 08:21, 29 April 2010 (UTC)


 * It is much easier for people who think they know what they are doing to add 'i' when they see "KB" which leads to people who like IEC to put the 'i' everywhere without thought or consulting the references to get the exact correct numerical value. This leads to mistakes so adding an 'i' and letting a bot convert to numbers is not a good choice because it promotes lazy editing and mistakes. Basically, if someone "cannot be bothered" to put in the exact number then don't let them be lazy and just add an 'i' in the hope a bot will come along to correct it. The guideline states to use the methods used in the references, not add whatever system someone has a personal preference for. For obvious reasons that is the best choice. The guideline also specifically states to use numbers when disambiguating so to try to claim the present wording can be "interpreted as a licence not to disambiguate" does not logically follow. To put my point another way, if someone "cannot be bothered" to disambiguate with the method that has consensus (numbers) then they should leave the article alone and allow someone who is bothered to disambiguate properly. Not all use of KB needs disambiguation because sometimes the surrounding text in the article makes it perfectly clear as to what is meant. Glider87 (talk) 09:27, 29 April 2010 (UTC)
 * You rebut a strawman argument. No one suggested haphazardly changing to the IEC units without checking whether the binary sense was intended or not, much less automatically via bot. No one suggested that switching to the IEC units would remove the need to provide the exact numerical quantity either. --Cyber cobra (talk) 02:57, 1 May 2010 (UTC)


 * To rebut your claim, Dondervogel suggested precisely what you claim wasn't suggested. i.e. He wrote "be carried out by a bot" and suggested poeple be allowed to add an "i" because they cannot be bothered to add a number. That kind of lazy editing is haphazard and doesn't involve rigourous editing standards, as I already explained above. Glider87 (talk) 23:22, 4 May 2010 (UTC)


 * To clarify what I mean above, not all use of KB needs disambiguation because sometimes the surrounding text in the article makes it perfectly clear as to what is meant. Or for example adding disambiguation unnecessarily encumbers the flow of the article where the approximate relative sizes (i.e. 2 KB compared to 4 KB, or 2 apples compared to 4 apples) is the important crux of the article's point and not the precise number of bytes. Also for example it is not normal to disambiguate a volume of liquid into the exact number of atoms or molecules. Sometimes adding too much superfluous information is not the best thing for an article.Glider87 (talk) 08:10, 30 April 2010 (UTC)


 * Exceedingly few—if any—major manufacturer of personal computers in the world, when communicating to their customer base, uses the IEC prefixes. Exceedingly few—if any—computer magazines directed to a general-interest readership use the IEC prefixes. Accordingly, the terminology is unfamiliar to our readership. It does no good if we provide links to the IEC binary prefixes to “teach” readers about these topics if the reader won’t encounter them in real life outside of Wikipedia. To use them would cause unnecessary confusion and this is a violation of the most basic principle of technical writing. History lesson: Wikipedia tried using the IEC prefixes for three years and the IEC prefixes still didn’t catch on in the world. I believe that is the definition, is it not, of insanity when someone tries doing precisely the same thing as what previously failed and expects a different outcome? (A word from Greg L’s counsel: Not that I am suggesting any editors here are insane; I am using a metaphor to illustrate the absurdity of expecting that Wikipedia will somehow change the world if we try again. Wikipedia isn’t that influential.) What we certainly need to do here is abandon this silly notion that just because there is a “standard” proposed by any standards organization (yes, including the BIPM and their SI), it is somehow incumbent upon organizations and publications to observe the standard if it does not enjoy widespread use in the real world. Wikipedia already ignores the rule of SI in order to not cause even the slightest bit of confusion in our readership. The rule of SI requires that a space always separates the number and a unit symbol and, per 5.3.7 Stating values of dimensionless quantities, or quantities of dimension one, this applies to the percent (%) symbol. Thus, it is supposed to be “Studies show that 5 % of wikipedians have no clue about proper technical writing” and not the “5%” required by MOSNUM. So, enough already about how we must follow every idea of the SI. Moreover, note that the above-described, rather minor issue of a space before the percent symbol wouldn’t cause galactic confusion for any reader; they would still correctly guess that “5 %” means five parts per hundred and would just assume it was a typo. However, writing “A typical USB thumb drive in 2010 had 8 GiB of capacity” would cause a *!* brain-interrupt and make many a reader to stop and go “Woa!”. We’ve been all through this before (ad nauseam). Nothing has changed in the world that requires us to revisit the issue again. Can you imagine what would happen if some poor fool read up on computers here on Wikipedia and marched into a computer store and said “I want a personal computer with Windows 7 and four gibibytes of RAM!” The customer would be met with A) confusion, B) a correction, C) knowing smirks, or D) laughter. Broaching the subject again here is just silliness by some who can’t let go, in my opinion. Greg L (talk) 02:03, 1 May 2010 (UTC)


 * P.S. “Gibibyte”, “Mebibyte”, and “Kibibyte” weren’t even in the Microsoft® Computer Dictionary (dictionary of computer terms) last time I looked at my local library. What the heck(??), readers can’t even look this stuff up elsewhere. Is Wikipedia to now get into some race to start using obscure language understood only by aliens and a few wikipedians? If that’s the objective, then to those who hold the BIPM as thy god of all that is good and ‘scientificy’ and want to start using words our readership doesn’t even recognize, I say “Hab SoSlI' Quch!” There: I out-obscured you. Greg L (talk) 02:41, 1 May 2010 (UTC)
 * I wouldn't exactly call Microsoft an authority on computer terminology. Also, I bet that publication is rather old. --Cyber cobra (talk) 02:49, 1 May 2010 (UTC)
 * Are you aware of another dictionary of computer terms that is as generally available at the libraries in English-speaking countries? The copy at my library is the latest, 2002 edition, which came out fully six years after the IEC advanced their proposal. None of them are there. Besides, this is an academic point, isn’t it? I’m sure you would agree Wikipedia would be bucking the trend observed by PC manufacturers and PC magazines when they communicate to a general-interest readership. Greg L (talk) 03:00, 1 May 2010 (UTC)
 * I wouldn't recommend using a dead-tree library resource in the first place. However, I concede the thrust of your point regarding prevalence. --Cyber cobra (talk) 21:19, 1 May 2010 (UTC)
 * Link to info regarding the previous 3-year IEC trial? --Cyber cobra (talk) 02:59, 1 May 2010 (UTC)
 * Sure. Scroll to the top of this page. Note the “archives” box. See the “Binary prefixes” section? We created an entire sub-classification for MOSNUM discussions dedicated to the ensuing binary jihad. Greg L (talk) 03:06, 1 May 2010 (UTC)
 * P.P.S. I once got a bit slipped into MOSNUM called “Follow Current Literature” (FCL). It got whittled away and eventually dissolved. But the principle is worth noting. The proper thing to do for any good encyclopedia is to follow the practices of current, reliable literature on any given subject so that readers can be properly prepared for their following studies elsewhere on the subject. Wikipedia is not a venue for editors to help smooth earth’s transition into the United Federation of Planets by adopting good ideas that simply aren’t used in the real world. “Thou shalt not cause unnecessary confusion in thy target readership” is a simple concept from Technical Writing 101. Arguments that we can’t precisely communicate binary quantities without the IEC prefixes simply don’t withstand the first pass of critical scrutiny. After all, computer magazines routinely manage to deal with the ambiguity all the time. Greg L (talk) 03:19, 1 May 2010 (UTC)


 * The Universal Serial Bus became a standard about the same time as the IEC binary prefixes (late 1990s). The example given above, "A typical USB thumb drive in 2010 had 8 GiB of capacity", would not require a footnote for USB to be understood by most Wikipedia readers. They may not know the technical differences between USB, FireWire, and  eSATA; but they have seen a USB device in the real world. -- SWTPC6800 (talk) 20:35, 1 May 2010 (UTC)
 * I doubt most users know that "gigabyte" etc. each have 2 meanings; having to explain the distinction is of similar awkwardness to explaining a new term ("gibibyte" et al.). --Cyber cobra (talk) 21:19, 1 May 2010 (UTC)
 * Most readers don’t need to know that there is a difference, which is why real-world publications seldom dwell on the subject. If the distinction between 10243 bytes and 109 bytes is important, then there are any number of ways to disambiguate without introducing terminology that is generally unfamiliar to our readership. It can be as simple as the footnotes used here in ''Power Macintosh 5500. Greg L (talk) 22:06, 1 May 2010 (UTC)
 * If the units are opaque enough to require explanation, I don't think the use of the IEC terms would add significantly to that opaqueness. The two senses of the units do indeed make quite a difference, there have even been lawsuits over it. In any case, ignoring IEC for the moment: since we are an educational publication, I think it would go against the mission to willfully gloss over [ignore] the difference between the senses, as you suggest we can/should. --Cyber cobra (talk) 00:28, 2 May 2010 (UTC)

(outdent) Fine. You have your opinion, Cybercobra and that is your right. I won’t waste my time trying to change it. I find it the height of absurdity to use precisely defined, unambiguous units such as “MiB” when they aren’t in the least bit familiar to our readership and aren’t used in the real world when communicating to a general-interest readership or customer base. To use them on Wikipedia would clearly result in even more confusion and lack of effective communication. Odd, don’t you think, that the rest of the computer magazine world manages to communicate about computer-related matters without using unfamiliar terminology? I can see that I’ve made every important point I can on this issue and there is no point further debating; positions have become clearer and seem to have hardened. Someone please give me a holler on my talk page if this goes to an RfC. We’ll make sure the RfC is (very) widely advertised and enjoys widespread input from all across our MOS, MOSNUM, and computer-article wikipedians. The vote count will likely be between 50 and 150 participants. Note too that there would have to be a clear consensus to change from the current practice, which is to simply follow the practices of current, reliable literature on the subject, which reflects the way the rest of the real world works when communicating to a general-interest readership. Then we can put this issue out of its misery… again. Goodbye. Greg L (talk) 01:17, 2 May 2010 (UTC)


 * These links may be of interest here.
 * http://www.onelook.com/?w=kibibit&ls=a and http://www.onelook.com/?w=kibibyte&ls=a
 * http://www.onelook.com/?w=mebibit&ls=a and http://www.onelook.com/?w=mebibyte&ls=a
 * http://www.onelook.com/?w=gibibit&ls=a and http://www.onelook.com/?w=gibibyte&ls=a
 * http://www.onelook.com/?w=tebibit&ls=a and http://www.onelook.com/?w=tebibyte&ls=a
 * -- Wavelength (talk) 21:27, 1 May 2010 (UTC)
 * There is also http://www.websters-online-dictionary.org/definition/kibibyte. -- Wavelength (talk) 21:44, 1 May 2010 (UTC)


 * What’s your point, Wavelength? No one said Web-based explanations don’t exist for these terms, which are real words due to the IEC’s proposal having languished for over 14 years. Probably only 0.02 centiuno of our readership is familiar with the terminology associated with this failed proposal. Some standard proposals gain traction in the real world; others don’t. It is not the job of Wikipedia to promote a standard that didn’t find traction in the real world any more than it is proper for us to write the unfamiliar-looking “75 %” (with the space between the number and symbol) just because that is the rule of SI. A small cabal was successful in sneaking the IEC prefixes into hundreds of Wikipedia’s articles nearly overnight, and it was like trying to cure a case of herpes to reverse that stunt after the fact. The odds are almost astronomically improbable that we’d ever change back. Greg L (talk) 22:06, 1 May 2010 (UTC)
 * My comments address the previously raised question "Are you aware of another dictionary of computer terms that is as generally available at the libraries in English-speaking countries?" I provided links to definitions in dictionaries which are not necessarily restricted to computing, and it seems to me that a validation by a general dictionary is better than one by a computing dictionary.
 * -- Wavelength (talk) 22:39, 1 May 2010 (UTC)


 * Ahhhh… I see. Indeed. One can find all sorts of words on the Web that, if used on Wikipedia, would result in logorrhea, such as antidisestablishmentarianism. Greg L (talk) 23:26, 1 May 2010 (UTC)


 * I very much doubt it will get to RfC stage because there is not a substantive new argument to justify it.Glider87 (talk) 09:28, 2 May 2010 (UTC)


 * Ignoring Greg_L's unhelpful rudeness (but requesting nonetheless that it not be repeated), it occurs to me that WP should aim to serve its readership better than popular computing magazines that provide ambiguous (and therefore confusing) information. (MOSNUM already requires unambiguous information, so the question is how to achieve this, not whether). The question I see is a simple one: Is one unfamiliar form of disambiguation (precise numerical value) preferred over another unfamiliar form (international standard prefixes)? There is no simple answer to this question, so I suggest a (polite) discussion here. Dondervogel 2 (talk) 16:13, 3 May 2010 (UTC)
 * Precisely my point. Nothing is lost (and there's a slight gain) by being technically correct in following the IEC standard when people are already unfamiliar with the exact value / dual meanings; inobvious/unclear is inobvious/unclear, so we might as well at least be technically correct since both approaches suck WRT clarity anyway. --<b style="color:#3773A5;">Cyber</b> cobra (talk) 18:11, 3 May 2010 (UTC)


 * Dondervogel 2 making baseless personally targeted accusations is a violation personal attack, in other words your statement is a personal attack. Greg is not being rude and he makes good points which you have not countered. Your argument is incorrect, using precise numerical value as disambiguation is not unfamiliar because it is the method used in the real world. IEC is not widely used in the real world and pushes a point of view that is incompatible with how Wikipedia operates. Therefore if any disambiguation is needed, it is not always needed, then using precise numerical value is much better than using IEC. Glider87 (talk) 22:41, 3 May 2010 (UTC)


 * Cybercobra, to counter your "nothing is lost" comment here are a couple examples of something "that is lost" by using IEC. The IEC system is not widely used and is unfamilar to readers so to use that system promtes a distorted view of the real world, which is contrary to WP:SOAP. To promote a distorted view of the real world is also contrary to WP:NPOV where articles must be written from a neutral point of view. This is a fundamental Wikimedia principle and cornerstone which cannot be lost. Glider87 (talk) 23:14, 3 May 2010 (UTC)


 * Well no, the words "cabal", "sneaking", "herpes", and "stunt" were a bit adversarial, though not as bad as a feedback loop where each countervolley is stronger than the volley that provoked it. Art LaPella (talk) 00:10, 4 May 2010 (UTC)


 * Greg has a point, I think the archives do show that IEC prefixes were quickly added by a small group pushing a minority point of view without much if any consensus. I think the archives also show attempts to change that were met with undue hostility. But that is all in the past. My point is that consensus isn't when a small group of people agree on a bad idea because a bad idea doesn't suddenly become good consensus just because a minority push it forward. Consensus also isn't when a small group with a minority point of view has a "polite discussion" and then agrees to implement that minority point of view. To keep on pushing a minority point of view when there is no substantive change to the topic can also violate guidelines like WP:GAME. Consensus comes from when a good idea is born, when people stop pushing their own non-Wikipedia point of view and use neutral logical rational debate. This is why we have WP:JUSTDONTLIKEIT and WP:ILIKEIT and WP:NPOV and others. This means anyone wanting to use IEC prefixes must first drop all notion of "I like it" and instead focus on how to reconcile their personal preference with the consensus existing in the cited guidelines and policies about neutrality. Although Headbomb is pro-IEC he also stated "When the computer industry decides to stop using K, M, G, etc... to mean powers of 1024, so will Wikipedia". This is a perfect example of leaving personal opinions aside and focusing on the logical rational consensus of existing policies. I cannot fault the logical basis for that statement because I think it follows the principles of all of the major Wikipedia policies about article writing. That is the "simple answer" Dondervogel is looking for. Glider87 (talk) 01:32, 4 May 2010 (UTC)


 * I would also suggest that when someone repeatedly violates WP:GAME and WP:DEADHORSE despite being asked not to then that is not polite or valid discussion. Glider87 (talk) 04:10, 4 May 2010 (UTC)


 * Glider87: I have responded to your accusations under Tony's Social Harmony section. Here I will focus on the substance of your argument, which seems to be that Wikipedia should follow the example of the computer industry.  But for as long as we strive for unambiguous statements, doing so is not an option here.  If there is a consensus on WP for deprecating IEC prefixes, then MOSNUM should deprecate IEC prefixes, and not otherwise.  I see no such consensus, either here or in the archives. Dondervogel 2 (talk) 15:47, 4 May 2010 (UTC)

Dondervogel 2: So, you think our readers will be familiar with terminology like “MiB” and “mebibyte”? Or, perhaps you agree that they are unfamiliar and *we’ll just teach ‘em.* Is that right? Even if we do teach them, do you think anyone will remember the terminology after they’ve been away from Wikipedia since it isn’t used anywhere else in the real world when communicating to a general-interest audience? Our adoption of this colossally failed standards proposal (it ranks right up there with the Uno and using a space before the % symbol) would amount to adopting our own “house lingo”, since ‘unfamiliar lingo’ that is endorsed by the IEC and the BIPM is still “unfamilar lingo’. This silly notion of yours was proven to be utterly bankrupt here and there is a snowballs chance in hell the practice would ever come back to Wikipedia. So why in the world do you persist here? Surely you must know there is no chance that what you desire would achieve consensus status here, do you not? Your persistence in the face of this reality is baffling to me. Greg L (talk) 21:38, 4 May 2010 (UTC)


 * Dondervogel, there is consensus for using familiar terms and not using IEC prefixes and it is stated in the guideline. Repeatedly claiming there isn't and linking long user talk pages doesn't make it so. To keep on claiming there is no consensus would be pointy, which can also lead to blocks. There is no hint or hope for consensus for what you want.  You, Dondervogel (Thunderbird2), were repeatedly asked to provide examples and you did not respond except to state personal dislike which according to WP:JUSTDONTLIKEIT is not a valid reason. This is another reason why claiming there isn't consensus fails to convince other editors.  The archives show you have been making these claims for years when it is obvious there is still no hint of IEC prefixes being the majority system in use so I suggest you stop beating the dead horse WP:DEADHORSE. Linking that talk page and drawing attention to the personal attacks contained on that page (where you again personalise an attack and accuse Greg of something) is also another violation of personal attack. Repeated violations of WP:NPA can also lead to blocks. Contrary to the conclusion on that talk page I think the reason why those people stay away from this page is because the issue has been debated fully and they have found the current consensus to be strongly valid. That is to say even if they personally don't agree with the conclusion they accept it as the logical Wikipedia thing to do. Like Headbomb who is pro-IEC but who also shows that Wikipedia policy is important. Glider87 (talk) 00:18, 5 May 2010 (UTC)
 * When the majority of the computer industry want to be unambiguous they use a precise number of bytes or use some form of power notation. Since Wikipedia follows the example in the real world then using precise numerical value or power notation is the logical choice as stated in the guideline. Can you can show how using a number like 1024 (apples/boxes/bytes/bears) is ambiguous? Glider87 (talk) 00:18, 5 May 2010 (UTC)

Is the party with whom I am speaking?

 * Dondervogel 2: I note that the entire contents of the current version of Dondervogel 2’s user page, which plays coy about the editor behind the account with the following:




 * If your user page is not intended to deceive and mislead (which would be an unfortunate violation of WP:Civility), then Dondervogel 2 and Thunderbird2 are one in the same. Assuming that to be the truth of the matter, then you, Dondervogel 2/Thunderbird2/whatever-you-wanna-call-yourself, should certainly remember the following post from an admin (∆ here), as follows:




 * Please note the “tendentious editing and forum shopping”-bit, above. It seems to be as applicable to Dondervogel 2 today as it was 15 months ago when you operated as Thunderbird2. Please just accept the community consensus on this matter. You haven’t brought up a single new thing that you didn’t mention last time around. Greg L (talk) 19:42, 5 May 2010 (UTC)


 * Ah, socks. Gotta love 'em. Headbomb {talk / contribs / physics / books} 20:10, 5 May 2010 (UTC)


 * I thought Dondervogel was a new editor to this; like the others. He elected to coyly allude to his identity on Dondervogel’s user page only two minutes after first posting here on WT:MOSNUM. Cybercobra was clearly new to all this; he asked for directions to all the previous binary jihad (the Binary prefixes section of the archives). If I had bothered to simply go to Dondervogel’s user page to see who the hell we were dealing with, the course of these threads would have taken a different tack and we could have wasted less time on this. Greg L (talk) 20:49, 5 May 2010 (UTC)

Reminder of our obligations to maintain social harmony at styleguides
Not to make a big deal of this, but I think we should take heed of ArbCom's view, recently reiterated, that the stability of styleguide pages and the maintenance of harmony on their talk pages are a serious consideration. I include myself in this respect. Tony  (talk)  09:11, 3 May 2010 (UTC)
 * I see that Glider87 repeats | his baseless accusation here, so I will reply here too. As it has more to do with social harmony than any other matter discussed on this page, I make use of Tony's new section: How on earth can asking an editor to refrain from rudeness be considered a personal attack? Dondervogel 2 (talk) 15:31, 4 May 2010 (UTC)
 * To exaggerate it, how could "Please refrain from arrogant stupidity" (strictly hypothetical, not directed at anyone here) be a personal attack? The Wikilawyering answer would be based on this quote from No personal attacks: "comments should not be personalized and should be directed at content and actions rather than people". But what bothers me more is that each of you has used the word "baseless" where "overblown" would be much more accurate. Can we get back to discussing this Manual of Style page? Art LaPella (talk) 19:24, 4 May 2010 (UTC)
 * Dondervogel, you accussed Greg of being rude when he is not rude and that is a personal attack because you personalised a comment against him. Note also in WP:NPA it is not a personal attack to state that your statement is a personal attack against someone else. If you study WP:NPA you will find the warning I posted is not baseless. Glider87 (talk) 23:04, 4 May 2010 (UTC)

The MOSNUM requirement for disambiguation
I'm starting a new thread because there's a point of general principle that requires clarification here. I have been assuming until now that there is a widespread consensus for the MOSNUM requirement to "Identify and define ambiguous units on their first use in an article". Is this assumption correct? Dondervogel 2 (talk) 16:37, 4 May 2010 (UTC)

My question was a general one and I think it is wise to steer away from the megabyte. Better examples for a general discussion might be the ton, mile and calorie. Under what circumstances should such units be disambiguated? Dondervogel 2 (talk) 06:22, 7 May 2010 (UTC)
 * Sure. When necessary. The rest of the computer magazine world gets along just fine by not being overly specific. A computer with “2 GB of RAM” is sufficiently clear and so too is a “1 TB hard drive.” If it is truly necessary to split hairs—perhaps because the topic covers fitting as many files as possible into available hard drive space—then it would be appropriate to be more exact. But one never does so by using terminology like “kibibyte”, which is unfamiliar to our readership. And why is that the case? Because it simply isn’t used by any computer manufacturer when communicating to their general-interest consumer base, nor by any computer magazine directed to a general-interest readership. It’s just that simple. There are many, many ways to “define ambiguous units” without setting Wikipedia backwards four years. One way is exemplified by this Power Macintosh 5500 article, which was an example of splitting hairs for precision that probably wasn’t really necessary, but was accomplished using terminology familiar to our readership. I find any reasoning based upon the logic that “we must identify and define ambiguous units on their first use in an article" = “we must use unfamiliar terminology like kibibyte” to be bankrupt non sequitur; the conclusion does not follow from the premise. I’m sorry, but you’ve been made aware of Wikipedia’s three-year-long trial using the IEC prefixes and that didn’t do diddley to promote the adoption of the colossally failed proposal by the IEC. All that resulted was that some of our articles made Wikipedia look foolish—as if we were so very anxious to speed Earth’s future membership into the United Federation of Planets. What you are suggesting simply has no chance in the world of passing here. Why do you persist? Greg L (talk) 21:53, 4 May 2010 (UTC)
 * They're unambiguous to computer people. The average person would likely think 1 GB of RAM and a 1 GB hard drive have the same space; and they'd be wrong, although in a surface sense it doesn't matter since a "1 GB hard drive" will have the same space no matter who they buy it from and cross-comparisons between RAM and HDDs are infrequent. Not that this [should] have/has any bearing on IEC (which I've now concluded is futile). --<b style="color:#3773A5;">Cyber</b> cobra (talk) 22:02, 4 May 2010 (UTC)
 * Fine. That’s why we have methods to disambiguate, like this Power Macintosh 5500 article, which doesn’t violate the foundation of Technical Writing 101: “Thou shalt not cause unnecessary confusion in thy readership.” Alleging that we can only address ambiguities by resorting to terminology unused in the real world is clearly false. Why do you persist? You must certainly realize that turning the clock back four years doesn’t have a snowball’s chance. Greg L (talk) 22:07, 4 May 2010 (UTC)
 * Did you not read the last sentence of my comment at all? There are of course other ways to disambiguate. --<b style="color:#3773A5;">Cyber</b> cobra (talk) 22:11, 4 May 2010 (UTC)
 * Other ways to disambiguate that use the minority unfamiliar IEC prefixes are not compatible with Wikipedia policy about how the example set in the real world should be followed. The way the majority of the computer industry disambiguates, when they need to, is to use a precise number or power notation. Both methods are already described in the guideline, both methods are unambiguous and both methods comply with Wikipedia policy about following the example set in the real world. If you can describe a way to disambiguate that does not use IEC prefixes then that might also be fine. Glider87 (talk) 23:29, 4 May 2010 (UTC)
 * I specifically gave up promoting IEC prefixes, you've completely misinterpreted my comment. "Other ways" would include giving the exact numerical value and similar. --<b style="color:#3773A5;">Cyber</b> cobra (talk) 01:01, 5 May 2010 (UTC)
 * I have not completely misinterpreted your comment. I stated that IEC is not compatible with Wikipedia because Dondervogel wants to try to use IEC prefixes. If you read the current guideline text the "other ways" you mention are already advocated. Glider87 (talk) 01:10, 5 May 2010 (UTC)
 * Yes, I was attempting to give an answer from a general perspective and disregard the (for all practical purposes concluded) present IEC debate. I was trying to be objective and reasonable; so sue me. --<b style="color:#3773A5;">Cyber</b> cobra (talk) 01:17, 5 May 2010 (UTC)
 * Cool. Glider87 (talk) 01:29, 5 May 2010 (UTC)


 * It already says what to do on the guideline page Disambiguation. Glider87 (talk) 07:42, 7 May 2010 (UTC)


 * Yeah, we know what you think, Dondervogel 2. But, as when you worked under the name, User:Thunderbird2, your contributions showed you to be a single-purpose account dedicated to the proposition that Wikipedia should continue to use IEC prefixes long after admins ruled there was a consensus on the matter. You don’t feel there was a consensus. As they say in the military: “So sad – too bad.” As can be seen on archive B9, extraordinary efforts were made to understand the basic viewpoints the editors had and achieve a consensus. Nothing has changed in the world’s circumstances to indicate we should have to repeat all that effort. So, to Dondervogel 2/Thunderbird2/whatever-you-wanna-call-yourself, you should certainly remember the following post from an admin (∆ here), as follows:




 * It would behoove you to note the “tendentious editing and forum shopping”-bit, above. It seems to be as applicable to Dondervogel 2 today as it was 15 months ago when you operated as Thunderbird2. Please just accept the community consensus on this matter. You haven’t brought up a single new thing that you didn’t mention last time around. Greg L (talk) 23:07, 7 May 2010 (UTC)

The IEC sub-topic from above

 * Regarding IEC from the parent section above: I think Dondervogel should now stop advocating IEC disambiguation on this talk page because the debate is for all practical purposes concluded, that is to say the consensus is IEC disambiguation should not be used on Wikipedia. Do you agree Cybercobra? Glider87 (talk) 01:29, 5 May 2010 (UTC)


 * I trust, Glider, that Cybercobra agrees. Judging from his latest post, it appears Dondervogel hasn’t yet let go of this. It’s time to put an end to this badgering. The larger Wikipedian community settled this issue long ago. There has been zero change in the world’s circumstances to warrant revisiting the issue. The community does not have to respond to Dondervogel’s refusal to Drop the stick and back away from the horse carcass. He should understand that the decision to adopt the present guidelines governing the expression of binary values was the subject of a huge RfC that ran for a long time. The Binary prefix archives were pointed out to him, which clearly showed that Wikipedia tried using the IEC prefixes for three years and that did nothing to change the fact that the world still soundly ignores the IEC’s proposal. He may not interpret others’ ignoring him from hereon as acquiescing to his wishes. He may not change MOSNUM unless there is a formal, widely advertised RfC on the subject in which there is a clear consensus to turn the clock back four years on this issue. Such an RfC outcome hasn’t a snowball’s chance in hell. As Tony pointed out above (∆ here), …the stability of styleguide pages and the maintenance of harmony on their talk pages are a serious consideration. Moreover, there is nothing to indicate that a repeat of the community-wide RfC is somehow indicated on this issue. Current MOSNUM guidelines on the expression of binary values are in keeping with the practices of every other fine encyclopedia, computer magazine directed to a general-interest readership, and PC manufacturer when communicating to computer buyers. Dondervogel may not circumvent the requirement for a full, Wikipedia-wide RfC by scraping together a couple other editors to tag-team in one of those absurd theatrics of “Oh my! Campy idea there Dondervogel. By all means. Let’s do!” The bar for defining “community consensus” is much higher than that on this issue. In short, the community is free to ignore him now. Moreover, if he persists, and won’t let go, we may have to deal with the matter as an issue of WP:Tendentiousness. Greg L (talk) 16:28, 5 May 2010 (UTC)


 * P.S. Since Dondervogel2 is apparently Thunderbird2, I guess the above post is nothing new to him. Greg L (talk) 20:55, 5 May 2010 (UTC)

Comment requested

 * 1) My edits are made in good faith
 * 2) I have requested comment here. Dondervogel 2 (talk) 16:47, 10 May 2010 (UTC)


 * As WP:AGF says "This guideline does not require that editors continue to assume good faith in the presence of contrary evidence." The contrary evidence to assuming good faith would be repeated attempts to ignore consensus, to push a point of view contrary to consensus and wikilawyering. As Gwen Gale says we know what you are trying to do, just drop the stick and stop beating the dead horse. Glider87 (talk) 23:40, 10 May 2010 (UTC)
 * I also think that Dondervogel2/Thunderbird2 should be topic banned from anything to do with binary prefixes. Glider87 (talk) 23:43, 10 May 2010 (UTC)


 * Indeed. That was a splendid demonstration of forum shopping that blew up in his face. Greg L (talk) 21:58, 11 May 2010 (UTC)