Talk:Binary prefix/Archive 18

Archive 15

Archive 16

Archive 17

Archive 18

Remove NPOV tag

FWIW I see little if any POV in the article. @Locke Cole: added the NPOV tag without initiating a talk discussion; in accordance with paragraph 4 of WP:WTRMT such a tag may be removed. Unless there is such a talk initiated in the next day or so identifying specific POV I will remove the tag. Tom94022 (talk) 06:24, 21 December 2022 (UTC)

Yup. I'd have to suggest that any reader coming across an NPOV tag on an article on this subject would be utterly baffled. Tags aren't a mark of shame, they are intended to initiate discussion about specific issues within an article. AndyTheGrump (talk) 06:54, 21 December 2022 (UTC)

@Tom94022: I've removed the template as the issue has, for the moment, been resolved. —Locke Cole • t • c 07:26, 21 December 2022 (UTC)

There is no consensus for this edit (or a similar one before) it so I reverted it. Specific problems with it include

• Confusing mention of metric prefixes in the lede of an article about binary prefixes.
• Change of name of one the columns to "JEDEC" without consensus.
• Listing of 'tera' as a JEDEC prefix (tera is not defined by JEDEC)

I can see the template being included further down the page, when binary prefixes are compared with metric ones.

Dondervogel 2 (talk) 08:42, 21 December 2022 (UTC)

I added the template at a more appropriate location, which resolves the first of the 3 bullets. The other remain unaddressed, but editors NebY and Headbomb discourage discussion of the JEDEC column. Perhaps they can explain why. Dondervogel 2 (talk) 09:00, 21 December 2022 (UTC)

I agree, there is no consensus... no consensus for removing the template that has been on this article for over ten years. Get consensus for that before removing it again. —Locke Cole • t • c 16:35, 21 December 2022 (UTC)

Pinging all who contributed to a discussion (about where to discuss the JEDEC header) at bits template @Headbomb: @Jc37: @Kbrose: @Locke Cole: @NebY: Dondervogel 2 (talk) 09:06, 21 December 2022 (UTC)

A POV tag is not appropriate just because IEC warriors refuse to use the template. Reverted to consensus version. Headbomb {t · c · p · b} 12:24, 21 December 2022 (UTC)

Cut the childish bullshit about "IEC warriors" (if you bothered to read the article you would see the template is used) and present a case for including decimal prefixes in the lede of an article about binary prefixes. Dondervogel 2 (talk) 12:51, 21 December 2022 (UTC)

Decimal prefixes are already in the lead. Second paragraph, even. As to IEC warriors, could you explain what someone is to think about the edit history at User:Thunderbird2/The case against deprecation of IEC prefixes (edit | talk | history | links | watch | logs)? According to the page statistics you've made over 400 edits as Dondervogel 2, and over 100 edits as Thunderbird2. You make around 20-40 edits to that page each year going back over a decade. Can you explain how someone seeing that wouldn't think you were pushing an agenda or a POV? —Locke Cole • t • c 16:41, 21 December 2022 (UTC)

Tom94022 and I worked on that essay. I didn't know that there was a rule against writing or maintaining an essay, nor against holding an opinion about the contents of an essay, whether for or against. Dondervogel 2 (talk) 17:34, 21 December 2022 (UTC)

If the intention here is to return to edit-warring, and to vacuous discussions about 'consensus' along with RfCs that weren't even notified on the talk pages they affected, it would seem to me that the only appropriate course of action is to return to ANI, and call for topic bans all round. I suspect the communities tolerance may be running thin.

Alternately, how about those involved each giving a simple explanation, without reference to 'consensus', RfCs etc, as to what exactly this dispute is about. An explanation that a reader of the article, looking for information on the subject rather than a long history of bickering about number prefixes etc can understand. Anyone incapable of providing such an explanation, in a short paragraph rather than reams of past internal Wikipedia history, should probably in the best interests of Wikipedia be told to go away, and find another damn hobby. And told to grow up. This vacuous round-in-circles nonsense is simply infantile, and achieves absolutely nothing of benefit to the article, to Wikipedia, or to the understanding of what is, after all, an argument about a relatively minor issue in an article on number prefixes, not the history of the Balkans. It isn't complicated. AndyTheGrump (talk) 14:23, 21 December 2022 (UTC)

The article is about binary prefixes, and if these are to be summarized in a table in the lead, this table should not include the column of decimal prefixes. Only the two forms of binary prefixes should be presented there. There is no call to transclude a template that is inappropriate in the context. —Quondum 14:40, 21 December 2022 (UTC)

You say 'should not', but can you explain why it shouldn't also include decimal prefixes, given that a proper understanding of the topic appears to be contingent on understanding that 'kilo', 'mega' etc are being used in two different ways in two different contexts? How does it not benefit the reader to give a simple illustration of what the prefixes mean in these two contexts - the decimal one they are most likely to be familiar with, and the binary one which they may not be? The purpose of this article is to increase understanding of the topic, and I cannot for the life of me see how imparting useful information could not be beneficial. AndyTheGrump (talk) 15:03, 21 December 2022 (UTC)

I agree the template is relevant background, which is why it is included in the history section. What is the benefit in including the template twice? Dondervogel 2 (talk) 15:21, 21 December 2022 (UTC)

Including similar tables twice may well be unnecessary. But if so, why not simply place the complete table at the top of the article, where readers new to the topic will see it, and immediately understand what the topic is about? AndyTheGrump (talk) 15:24, 21 December 2022 (UTC)

As the article also references terms that many readers are familiar with (and that the wide world that is our source still uses dominantly even in new uses) it makes sense to present the kilo/mega/giga terms side-by-side with the "newer" terms. The lead of our article even has this paragraph:

In most contexts, industry uses the multipliers kilo (k), mega (M), giga (G), etc., in a manner consistent with their meaning in the International System of Units (SI), namely as powers of 1000. For example, a 500-gigabyte hard disk holds 500000000000 bytes, and a 1 Gbit/s (gigabit per second) Ethernet connection transfers data at nominal speed of 1000000000 bit/s. In contrast with the binary prefix usage, this use is described as a decimal prefix, as 1000 is a power of 10 (10³).

which directly references the decimal names/meaning, so the table being present in the lead (in the form of an invocation of the template) makes sense so our readers can see the terms side-by-side and quickly understand the difference. —Locke Cole • t • c 16:02, 21 December 2022 (UTC)

I understand your point but I disagree. The article is about binary prefixes and these should be introduced first, without the complication of decimal prefixes. The comparison should come second. Dondervogel 2 (talk) 19:21, 21 December 2022 (UTC)

AndyTheGrump, sad to say, you do not seem to have been paying attention. You lost control of this bunch from the start. I answered your question as you asked about my perspective, succinctly and IMO sufficiently. What I refrained from saying was that it got hijacked into an old (and unrelated) fight; none of what followed was "about" or was stated to be to restore the column, only an aggrieved insistence on re-including the controversial template. —Quondum 21:28, 21 December 2022 (UTC)

I have yet to see any rational explanation of why a template illustrating the topic of this article is 'controversial'. Or, more to the point, why such controversy is of any relevance to the readers of this article. The fact that people keep bringing it up suggests to me that it might be better to find new contributors to work on it - contributors who aren't apparently obsessed with facile arguments serving no useful purpose. If there was politics involved here, or religion, or nationalism. or even support for your local soccer/baseball/whatever teams, I could understand it, but people seem to be arguing just for the sake of it. AndyTheGrump (talk) 23:00, 21 December 2022 (UTC)

Of course, but the objection to introducing these binary prefixes for unambiguous units for storage has indeed been a religion to a group of vocal and angry editors who find no effort too great to prevent their promulgation or even admission of benefit, when every standards body in the world promotes them, declares old usages deprecated, and a growing body of software vendors implements them to remove the confusion that has persisted throughout the growth of information technology. So, this full table is useful to these refuseniks, because of the last column (JEDEC), to hang on and continue the gaslighting effort in making new readers believe that at least one body (JEDEC) still supports or defines the old usage (which JEDEC does not, in clear language). The opponents however want to remove the table, because they can't seem to get enough courage or support to just remove the misleading last column. This would not be an issue in this trivial situation, if the table did not contain that last column. kbrose (talk) 00:02, 22 December 2022 (UTC)

The 'old usage' may well be deprecated, but isn't it something we'd expect at least some of our readers to have encountered? I don't think pretending that prefixes haven't been used that way is helpful to the understanding of the topic. AndyTheGrump (talk) 00:21, 22 December 2022 (UTC)

Are you under the impression that anyone is trying to remove information about the "old" units, e.g. by mentioning that most standards bodies deprecate them? Is there something wrong with presenting it all in a balanced way? Your earlier characterization of, effectively, religion or sport fanatics is pretty close. BTW, by "cointrovertial" I meant "contested", and meant the template on its own. —Quondum 00:48, 22 December 2022 (UTC)

Certainly the old usage is still common and it is not suppressed anywhere. This is indeed discussed in the many articles about the subject matter, and I have observed not a single editor here to just deny that process. These units are often explained and contrasted. It is often documented in great detail. Everyone agrees on that. It is just one side of the issue who does not want to acknowledge the transition to new standards. Wikipedia attracts new writers all the time, and often, if not commonly, they want to use new units because they are in fact commonly used in much new software these days, in the most widely distributed operating systems, even. But the soldiers of denial religiously revert the new editor's work to the old ambiguous usage. The censorship is blatant, anti-intellectual, and unencyclopedic. When two new unit prefixes where introduced this fall by the SI, all Wikipedia editor were eager and quick to comply and started using them immediately within days, a courtesy not conveyed to the binary prefixes by that group of editors. These are even less know than the binary prefixes. A common (fake) objection always involves that the binary prefixes are not known by most readers. Well, if readers knew everything they would not read Wikipedia. But denying new units certainly does not help them learn new things. There is rarely ever anything that is not explained on WP. kbrose (talk) 01:10, 22 December 2022 (UTC)

Well I didn't expect to set off such a s**tstorm, but now having read thru the above I do suggest that if the JEDEC column is to remain within the template in the lede, then the lede should also have some mention of what JEDEC is, since I suspect it is not a term familiar to the average reader. Perhaps just appending to the 3rd paragraph of the lede something like ... these three usages are referred to as a binary prefixes and have been standardized by the Joint Electron Device Engineering Council (JEDEC). Note I have spelled out JEDEC as is already done in the lede for SI and IEC. I would also suggest then for consistency we move the links to IEC, SI and JEDEC out of the template.Tom94022 (talk) 00:51, 22 December 2022 (UTC)

Exactly. Only engineers know who or what JEDEC is. Yet, the intent of the column is even obscure, because it lists only three or four prefixes, and the 'old usage' does extend beyond those. In additions, JEDEC makes no claim to these 'definition', they acknowledge the deprecation, and refer to the other standards bodies that have been involved in units. JEDEC does not define or recommend any. They just use the units. Microsoft is the single most visible supported of the old unit usage, and the column could be filled fare deeper based on actual usage. Yet, the column is labeled 'JEDEC'. It should be labeled 'Deprecated' and people would understand it and it would be correct. kbrose (talk) 01:19, 22 December 2022 (UTC)

I support the change to the 3rd paragraph suggested by Tom94022. I'm not sure about removing the links from the template. They made be needed for other uses of the template. Dondervogel 2 (talk) 01:14, 22 December 2022 (UTC)

Is there any particular reason why we need to use a template at all? Exactly the same information can be conveyed in a custom-made table. AndyTheGrump (talk) 01:18, 22 December 2022 (UTC)

Yes, a local table could be made, but that just increases maintenance effort if the table is used in multiple places. There is also a reluctance by some editors to discuss the JEDEC column at all. IMO it's a better use of editor's time to gain consensus once and then apply the template where it is relevant. Dondervogel 2 (talk) 01:29, 22 December 2022 (UTC)

The last attempt to 'gain consensus' didn't seem to work very well. And in any case, templates are supposed to be a means to aid article creation - they aren't supposed to be used to enforce common content where needs may differ. AndyTheGrump (talk) 01:39, 22 December 2022 (UTC)

Dondervogel 2, I tend to disagree with your assessment, in the current context. These templates have wasted many orders of magnitude more effort and time than individual table creation per article. It is the rare table that makes sense to templatize, because tables are so context-sensitive. And until a template is actually no longer contended, your argument makes no sense.

BTW, as a side-comment, use of a template in the MoS has served as a strategy to allow a small group of editors changes of MoS in without debate in the MoS forum. Use of a contested template to dictate MoS is perverse. —Quondum 02:24, 22 December 2022 (UTC)

That is exactly why the JEDEC column should be removed, because its content is so specific to a single vendor and has been used to influence opinion just by the type header it is given. Table templates should contain only indisputable facts that cannot be changed or reinterpreted. That's the nature of the other columns in that table. Any deviation from these true standards should be addressed in each article. kbrose (talk) 03:21, 22 December 2022 (UTC)

Using 'K', 'M' etc in the manner described in that column isn't 'specific to a single vendor'. It is a common usage. AndyTheGrump (talk) 03:28, 22 December 2022 (UTC)

It's not only common usage, it's the dominant, near universal usage. Headbomb {t · c · p · b} 04:01, 22 December 2022 (UTC)

• @AndyTheGrump: The last attempt to find consensus failed because the RfC was ill-conceived. There were too many questions posed at once and the questions were posed without prior discussion. There were too many options, an option favoured by some editors was omitted, and the responses to which depended on the selected outcome of another. It was also peppered by childish contributions from editors bringing up issues that were unrelated to the RfC. I tried unsuccessfully to simplify the RfC by removing some of the options that seemed least likely to succeed, so we could focus on the others.
• The failure to achieve consensus does not make the problem go away. Quondum's view is that the templates are not useful, presumably with the intention that individual articles make their own tables, and their own choices. My point is that this multiplies the number of debates by the number of pages. If that is the consensus, so be it. Is it?
• The question I wanted to settle (after the RfC closed) was where to discuss a content dispute at 'bits and bytes'. Should it be at the talk page of that article or where the discussion on the 3 templates has been centralised for over a year. Several think it is better not to discuss where to debate, and yet here they are debating the same issues on yet another page.
• For pages that choose to go their own way, there is no need to centralise the debate. For pages that prefer to use the template, shouldn't there be a centralised location to discuss those templates?

Dondervogel 2 (talk) 08:39, 22 December 2022 (UTC)

Back to the subject article

A template is not policy. A template is not a guideline. A template is not a constituent part of a manual of style. Discussions concerning the content of templates cannot be used to resolve content disputes, because their use cannot be imposed on any particular article. They are a tool for making content creation easier, and that is all they are for - any attempt to use them to enforce particular content, or a particular presentation of content, is an abuse of process. Forget about the damn template, and stick to explaining how adding or excluding specific material in this article aids reader understanding of it. AndyTheGrump (talk) 09:02, 22 December 2022 (UTC)

About Binary prefix then:

• This article is about binary prefixes, so it should start by explaining what is meant by a binary prefix. Binary prefixes are related to decimal prefixes, so it makes sense to explain how they are related. That should come second.
• The "JEDEC" column is part of explaining binary prefixes, but the name is contentious, and the content of the column depends on the choice of name. The first question is what to call it. Reasonable possibilities for the column header IMO include

1. Computer memory (they are used mostly for computer memory)
2. Deprecated (because their use is deprecated universally by international standards)
3. JEDEC (because JEDEC defines the first three)
4. Legacy (the compromise reached at 'bits and bytes' Quantities of bits)

• Unacceptable options that have been proposed include

1. Common (because the decimal interpretation is also common)
2. Computing (because the decimal interpretation is also widely used in computing, for communications and data storage)
3. Customary (because the decimal interpretation was used long before the binary interpretation)
4. Traditional (because the decimal interpretation is also traditional)

Dondervogel 2 (talk) 09:55, 22 December 2022 (UTC)

Please can you clarify "Legacy (the compromise reached at 'bits and bytes')"? I haven't found such a compromise prevailing in the history of Template:Bit and byte prefixes or being reached at Template talk:Bit and byte prefixes. Our Bits and bytes article is about a TV series. Do you mean somewhere else? NebY (talk) 10:52, 22 December 2022 (UTC)

The discussion was held at Quantities of bits. The compromise was reached in November 2021 and implemented in this edit at bits (and similar edits at the 2 sister templates). It lasted about 10 months, until this edit in September 2022 upturned the apple cart, resulting in the present instability. Dondervogel 2 (talk) 11:13, 22 December 2022 (UTC)

Reality: The templates were being actively discussed for many months with no consensus being reached, when you added another proposal during a period of inactivity that was hastily agreed to in less than a week without notifying any of the other editors involved. After said edit was finally noticed many months later it was undone because of the aforementioned lack of notification to other discussion participants. You don't just get to argue indefinitely until you wear your opponents down or catch them missing discussion and get to call that "consensus". I do wish you'd stop casting aspersions and flat out lying about events. —Locke Cole • t • c 16:40, 22 December 2022 (UTC)

You are correct. The discussion was not at 'bits and bytes' (by which I meant Bit and byte prefixes, but it was not there either). I have now struck out and corrected the misleading text. Apologies for the confusion. Dondervogel 2 (talk) 11:20, 22 December 2022 (UTC)

AndyTheGrump, just a note on "A template is not a constituent part of a manual of style." Please take a look at MOS:COMPUNITS: {{Bit and byte prefixes}} is transcluded there. Its presence there has been used to argue that it is a constituent part of the MoS. —Quondum 13:40, 22 December 2022 (UTC)

And look where such arguments have got us. A poorly-notified, badly-worded, inconclusive RfC on a template transcluded to a MOS sub-page cannot be used to determine content. The MOS itself cannot be used as such. It is a guideline - one that states, at the top of the page (as with all MOS pages) that it is "best treated with common sense, and occasional exceptions may apply". The purpose of a manual of style (anywhere, not just on Wikipedia) is to provide guidance on how subjects are written about. They are not rules restricting content. Using them as such is an abuse of process. AndyTheGrump (talk) 15:01, 22 December 2022 (UTC)

– agreed. My point is of-topic for this talk page, of course, unless of course someone actually makes an argument such as that this page should look like the MoS, which transcludes the template, or that the template is a guideline because it is part of the MoS. But it should not be transcluded in the MoS, IMO. But I do not venture into that crowd. —Quondum 15:29, 22 December 2022 (UTC)

Agreed that it shouldn't be transcluded into the MOS, because it is unhelpful to have MOS text saying one thing beside a table which sometimes contradicts it, depending on how the template's been edited. I've recently proposed that but it's met with some opposition. NebY (talk) 15:46, 22 December 2022 (UTC)

Ah, I see, indeed. This is all a bit too surreal for me. I have been caught up in a bigger picture that I've not been following, including people discussing me behind my back in a forum that I have not been watching. —Quondum 16:04, 22 December 2022 (UTC)

Just so. Contrariwise, my interest has been in the impact of template changes on the MOS, but now I get pinged to places like this. NebY (talk) 16:17, 22 December 2022 (UTC)

OK, so I accept responsibility for both sins (opposition at MOS and pinging you here). Here is my position on both

• If there is consensus for creating a bespoke table at MOS I would happily abide by it, but I see no such consensus. My opposition is based on the notion that it is more efficient (better use of volunteer editors' time) to hold one centralised discussion than multiple parallel ones.
• I pinged you here because you consider the content of the JEDEC column unworthy of discussion, yet here we are discussing it. You have not explained why it's OK to discuss the JEDEC column here but not there.

Dondervogel 2 (talk) 16:29, 22 December 2022 (UTC)

How about we stop discussing templates and the MOS entirely, and stick to discussing how best to make this article useful to readers? AndyTheGrump (talk) 16:38, 22 December 2022 (UTC)

Sure. I've lost track of who's for or against what in this article anyway, so I'll just suggest

• it's useful to have a table in the lead
• the first thing the table should do is show the two sets of symbols for binary prefixes, together with their values
• the table should then show the decimal series for comparison and contrast. It's helpful to our readers to provide that grounding and be clear from the start about the difference between binary prefixes, the article's subject, and the decimal prefixes with which most readers will be much more familiar.

NebY (talk) 17:25, 22 December 2022 (UTC)

That is precisely my position. Dondervogel 2 (talk) 17:47, 22 December 2022 (UTC)

That would seem sensible. AndyTheGrump (talk) 18:16, 22 December 2022 (UTC)

I agree with the statements listed. —Locke Cole • t • c 15:18, 23 December 2022 (UTC)

Me too Tom94022 (talk) 18:54, 23 December 2022 (UTC)

FWIW, this template is displayed in relatively few articles, most importantly, this one, Metric prefix, Unit prefix and MOS. After the current edit to this article only, the Unit prefix article lacks descriptive language for JEDEC in the text of the section displaying the template. The Metric prefix text is

In some fields of information technology, it has been common to designate non-decimal multiples based on powers of 1024, rather than 1000, for some SI prefixes (kilo-, mega-, giga-), contrary to the definitions in the International System of Units (SI). The SI does not permit the metric prefixes to be used in this conflicting sense.[29] This practice was once sanctioned by some industry associations, including JEDEC.

which i find rather compelling. Perhaps we should end this discussion here move to the template talk and work on a set of suggested text associated with template explaining the three column headings SI, IEC and JEDEC. If we wish to change JEDEC to something else then the only one I see as reasonable is "Customary" which would be an invented Wikipedia word and it along with IEC and SI would have associated text in every article in which the template appears. Tom94022 (talk) 18:54, 23 December 2022 (UTC)

This is the talk page for the 'Binary prefix' article. Whether or not we use a template defined elsewhere will be determined by consensus. Here. If you want to engage in discussions regarding a template we are under no obligation to even use, please do so elsewhere. AndyTheGrump (talk) 19:02, 23 December 2022 (UTC)

It seems there is consensus to use the template in this article but at least on editor has issues with the label on one column and maybe those issues may have been resolved by the text added to the lede. So yes we should end this discussion and move any issues with the one column to the template article. If a change is agreed upon there then we can revise all the articles linked. I do think each non-obvious template column heading should be summarized within any using article as in now the case herein - hopefully there is now consensus on that too. Tom94022 (talk) 19:28, 23 December 2022 (UTC)

I can see precisely zero evidence of any such consensus. NebY's proposal above has the support of four different contributors. Nobody so far has opposed it. NebY's proposal refers to a table specific to this article. Not a template. AndyTheGrump (talk) 19:43, 23 December 2022 (UTC)

A template is a means to an end. The three items NebY listed are all met via the template. There's argument over the details, but there's no need to try to disparage the template. Whether it's raw wikitext tables in the article or the template translucion the core bullet points are addressed. —Locke Cole • t • c 17:27, 24 December 2022 (UTC)

For the purposes of this article, the items three wrong order the template are in. I propose a different order, and that requires a table made for this article. As no-one's said no yet, I'll further suggest

• for binary units, we should show unit names/prefixes, then unit symbols, and finally unit values
• then for decimal units, we should show unit names/prefixes, then symbols, then values

This is the same approach we use in tabulating SI prefixes and many, though not all, of our articles about units. NebY (talk) 17:50, 24 December 2022 (UTC)

I give up

I had a vague hope that as an outsider to this article, I might be able to offer something new to the discussion, and in particular to discourage the endless round-in-circles bickering that seems to have become the norm, both here and on related subjects. Clearly I was mistaken, since it appears that many contributors here - quite possibly the majority - are more interested in engaging in endless cycles of vacuous juvenile POV-pushing over the supposed merits of alternative ways of representing numbers than actually fulfilling the supposed objective on an encyclopaedia: to provide an overview of what would seem to be a simple enough topic for readers with varying degrees of prior knowledge.

My only future involvement in this topic is likely to be in calling for contributors who have acted in such a manner to be topic banned and/or blocked from editing entirely. Wasting your own time is fine, but find somewhere else to do it, where you don't waste other peoples'. Readers deserve better than this. AndyTheGrump (talk) 18:25, 24 December 2022 (UTC)

Yup, as someone who has only been involved in this for a couple of years, it is certainly discouraging seeing some of the arguments used here. For something that should be as simple as just following our reliable sources, it seems to always get turned into what a small fraction of editors "like". Sorry to see you go so soon. —Locke Cole • t • c 18:32, 24 December 2022 (UTC)

If you are under the impression that I have found your contributions to this discussion of net benefit, you are mistaken. AndyTheGrump (talk) 18:34, 24 December 2022 (UTC)

Oh, I'm aware. I can still appreciate an uninvolved voice trying to wade into this issue and be sad to see you go, no? —Locke Cole • t • c 18:40, 24 December 2022 (UTC)

I concur with AndyTheGrump. This has been a futile endeavor for years. A small group of 'editors', the refuseniks of unambiguous units, always put up a kind of militant fight against fair representation, because they simply do not accept usage of the IEC prefixes under any circumstances. Motivational, but unintelligible ramblings like User:Locke Cole/IEC units are bad are examples that the opposition has no logical, encyclopedic goal, and that user's escalating of the polarization and bullying by simply removing opposing talk page contributions is exemplary (And now claiming the right to accuse me of edit warring). These are the methods that have governed the binary prefix discussions for years. Several times I have pointed out that this recursive history will never lead anywhere, because there is no real logic involved, it has become a matter of religion. And now the innocent claim of being an "uninvolved voice", when this user is at the center of the issue, a constant editing warrior against anything that agrees with the standard orgs. Benevolent editors come, try something, and go after seeing this rubbish unfold every time, and they leave because there are more productive aspect of life to explore than battling these people. And so the refuseniks keep winning because they won't go away so easily. That's how the MOSNUM got created too, in support from recognized sock puppets. For long periods of time I tried to stay away, still try to stay away every day, because it is just unrewarding, but every so often a particularly aggressive naysayer, shows up, trying to reverse the increasing acceptance of IEC prefixes here and everywhere. kbrose (talk) 20:45, 24 December 2022 (UTC)

JEDEC has not standardized any prefixes or units

In an edit today in the lede of the article it was asserted that JEDEC "has standardized the binary interpretation", which is a statement completely unfounded. JEDEC does not standardize any unit, and explicitly states that the binary interpretation is deprecated, pointing to the proper standards bodies for reference. They explicitly state that the units are only LISTED because of traditional practice. kbrose (talk) 21:14, 22 December 2022 (UTC)

You can quibble about the wording but it remains a fact that JEDEC defines K, M and G as 1024, 1024^2 and 1024^3. It is also a fact that it does not define higher order prefixes. The contested edit should be reinstated IMO. Dondervogel 2 (talk) 22:04, 22 December 2022 (UTC)

JEDEC lists the usage, per their own explanation, this does not constitute a definition, just an explanation of common usage, and that without references, btw. Why do you insist on the phrasing define. Going further JEDEC does not standardize, they do the opposite here; they deprecate the usage. Clearly they stopped updating the listings, because they had already deprecated their usage and the higher ones were not common. So there is no justification for that phrasing. And why are these limited JEDEC listings so important to you, to anyone, When the common behavior is to use these unit meaning beyond G? Why don't you want to list the rest? There are certainly references for that that are far more meaningful than the JEDEC papers. kbrose (talk) 22:43, 22 December 2022 (UTC)

• I use the word "define" because JEDEC defines kilo (K) (as a prefix to units of semiconductor storage capacity) as A multiplier equal to 1024 (210), and similarly for mega and giga. That is not a list.
• JEDEC is not important, but it is what the template currently uses. Acceptable alternatives include 'Computer memory', 'Deprecated' and 'Legacy'. Use of any one of those would mean we could populate the table to higher orders, which would IMO be more useful for the reader.

Dondervogel 2 (talk) 23:30, 22 December 2022 (UTC)

'Deprecated' and 'Legacy' are not acceptable because they are neither deprecated by the industry, nor are they legacy prefixes. Customary would be fine however if we want to expand the table to cover more than just the JEDEC entries. Headbomb {t · c · p · b} 01:43, 23 December 2022 (UTC)

• You might not like it but Deprecated is accurate (the JEDEC use is deprecated universally by international standards - that's what the word means) and Legacy is accurate (the JEDEC use remains the norm - that's what the word means). In fact Legacy was used (and stable) for a period of 10 months.
• Computer memory is an option if the table includes decimal prefixes, but would not work in a reduced table, with only binary interpretations.
• What is the objection to JEDEC (stable for years, before the consensus changed to Legacy)?

Dondervogel 2 (talk) 09:34, 23 December 2022 (UTC)

The word deprecated does not mean deprecated by standards body no one listens to, it means deprecated IRL. And IRL, they aren't deprecated at all. Legacy is also not accurate, because these aren't legacy units at all. They are the modern units everyone save an extreme minority use. Headbomb {t · c · p · b} 10:34, 23 December 2022 (UTC)

You might not like it but Deprecated is accurate (the JEDEC use is deprecated universally by international standards - that's what the word means) The problem is that a standards body deprecating something when the real world does not puts us in a position of using Wikipedia's voice to say something is Deprecated when it is still in regular and wide use. "Deprecated" (note the quotes) would be the more accurate shorthand, if and only if we wanted to give these international standards bodies that much weight (a question for WP:UNDUE I'd say considering, again, how rarely used the "standards" are in this instance). What is the objection to JEDEC It incorrectly implies only RAM/memory manufacturers use the terms, when many other industries still do as well. —Locke Cole • t • c 15:18, 23 December 2022 (UTC)

It might be advisable at this point to avoid getting into the minutia of details as to what exactly JEDEC has or hasn't specified until we decide the broader issues regarding the disputed table. AndyTheGrump (talk) 22:08, 22 December 2022 (UTC)

As long as JEDEC appears in the template it should be described in the lede so I am going to restore it with the word "defines" instead of "standardized." Details such it was by JEDEC deprecated can be in the body. Tom94022 (talk) 22:44, 22 December 2022 (UTC)

Maybe the lede now has it backwards - didn't JEDEC just define the terms as used first by semiconductor memory and then picked up by the OSes? So maybe we should resequence the sentence and lead with JEDEC?Tom94022 (talk) 22:50, 22 December 2022 (UTC)

This is not an article about JEDEC. The prefixes in question were in use before the standard concerned was created. Can we please not get involved in yet another discussion about matters not directly related to the article subject itself. Which is binary prefixes themselves, rather than one particular standard that describes them. Sort out the table question first, and maybe we'll get somewhere. AndyTheGrump (talk) 23:43, 22 December 2022 (UTC)

To solve the issue, I would just remove the JEDEC column altogether. It is redundant. The template is named "Prefixes for multiples of bits (bit) or bytes (B)", not "All prefixes for multiples of bits (bit) or bytes (B)" - as discussed in the text, there were more nomenclature/prefix proposals over time, and if we were heading for completeness in a comparison table we'd have to list them as well - no. Seeing this from a pragmatic point of view, we would not even loose any information as the four entries in the JEDEC column are already present on the left side of the table. Yes, it's labelled "Decimal" there, but everyone who uses kilo to mean 1024 knows that this is "jargon" (though common and convenient) and that the SI strictly defines kilo to mean 1000 only (and never did differently). It's not as if there would be a genuine proposal of some alternative nomenclature to the IEC prefixes which we would have to list in the table - before the IEC prefixes there was just ad-hoc (ab)use of existing nomenclature - and this is discussed in the article already. So, someone looking up the table for a quick check of what was the next prefix in the row after giga, or tera, will still find the info and simply apply the binary meaning to the decimal prefix, like before. Adding the same info in a JEDEC column does not add anything, but create the invalid assumption JEDEC would (have) endorse(d) the usage of the decimal prefixes for binary usage, which they clearly do not (and never did). All the nitty-gritty details regarding JEDEC are (and should) still (be) explained in the body of the article. In addition to solving the issue, this would also remove the somewhat artificial break after tera, make the table somewhat narrower (better flow on mobile devices) and add some nice symmetry to the table, which makes it even easier to "grasp" without having to "decode" the meaning.

--Matthiaspaul (talk) 10:57, 24 December 2022 (UTC)

Exactly. I have been advocating for this for a long time. The column makes no sense, is factually wrong and misleading. kbrose (talk) 17:46, 24 December 2022 (UTC)

Unfortunately for binary prefixes the IEC column is the outlier, presenting it in a table like this gives it undue weight on its own considering the overwhelming majority of our sources still use common terms like megabyte, gigabyte, etc. For a detailed look at what our sources say, see this page. —Locke Cole • t • c 17:30, 24 December 2022 (UTC)

Stop the utter nonsense. You are blind because you are prejudiced from the start. The metric usage of the SI prefixes has been the dominant use for decades, outside of some storage, and the IEC prefixes are an international standard now, and are the preferred units for operating system programming in the most widely distributed systems. Today any computer professional or just enthusiast encounters IEC prefixes in use almost hourly in different applications. The standards-correct usage of units is almost pervasive in new software and with open-source software we can actually evaluate this. kbrose (talk) 17:46, 24 December 2022 (UTC)

Odd, I haven't seen them in roughly 2 years in any software, and I modified that script to display download speeds in MB/s instead of MiB so it matched every other piece of software I use (also by the definition of MiB it was using it wrong, although I don't like it so I would have deleted it anyway).

the preferred units for operating system programming in the most widely distributed systems.

Yes, units, the things you famously need to program operating systems in.A Shortfall Of Gravitas (talk) 08:49, 10 January 2023 (UTC)

Add examples of IEC prefix in other units

The article say "IEC proposed kibi, mebi, gibi and tebi, with the symbols Ki, Mi, Gi and Ti respectively", so, let's add some few examples

• 1024 meters is 1 KiM (not lower "kim")
• 1024² meters is 1 MiM
• ...

Not popular, not used yet (the geographical square grids of ISO 19170-1:2021 need it), but a binary prefix is for it. It is the intention of the binary prefixes. Krauss (talk) 23:26, 30 April 2023 (UTC)

Although the prefixes are intended for any unit, it would not be in the way you suggest. The only example I can think of in the wild (other than for bit or byte) is the kibihertz (KiHz). For the metre it would be Kim (for kibimetre) and Mim (for mebimetre). Dondervogel 2 (talk) 07:58, 1 May 2023 (UTC)

Ok (!), KiHz is a good example, and we have some appeal in the ISO standards, ISO 266:1997 say explicitally: "Another series of frequencies that is in use is based on the definition of the octave as the frequency ratio 1:2. The frequencies of this series are calculated as powers of two (IEC 1260 base-two series).".
PS: IEC 1260 was superseded by IEC-61260-1.

Excessive but pointless examples

There have been literally millions of documents (manuals, articles, spec sheets, etc.) that used decimal prefixes to mean binary ones. The examples cited in the article are at the same time too few and too many. The text should be heavily summarized, and the references should be pruned to the most notable references, that are not dead links and have not been modified. Jorge Stolfi (talk) 17:39, 11 August 2023 (UTC)

Bad references

Many of the <ref> entries are dead, or are webpages that have been updated since the last access. Many of those entries also lack the date of the document (as opposed to the date of access). Also many of those entries are not source references but footnotes; these must be merged into the text or eliminated. I will try to do some cleanup, but it is a huge task. Jorge Stolfi (talk) 17:42, 11 August 2023 (UTC)

Submultiples?

I'd assume they would be called queci (qi), roni (ri), yoci (yi), zepi (zi), atbi (ai), febi (fi), pibi (pi), nabi (ni), mici (ui or μi), mibi (mi), cebi (ci) & debi (di) + dabi (Di) & hebi (hi). 2600:4040:208F:2A00:9CD:1511:4A28:A33D (talk) 03:09, 5 November 2023 (UTC)

All good guesses, but no, submultiples have not been defined. Dondervogel 2 (talk) 20:19, 5 November 2023 (UTC)

ISO is more recognizable than IEC

I propose that all references in the article and the tables to IEC binary prefixes should be modified to refer to ISO binary prefixes or ISO/IEC binary prefixes. ISO is simply more recognizable than IEC. Z80Spectrum (talk) 18:49, 28 January 2024 (UTC)