Talk:Decimal separator

Seems like SABS M33a is the correct source for the information. I do not have a copy of the standard.....

See https://www.sabs.co.za/Business_Units/Standards_SA/WebStore/search/detail.aspx?id=4226&lang=EN MoHaG (talk) 04:31, 26 November 2009 (UTC)

That link is quite dead.

I have never seen a comma used as a decimal point in South Africa and I've lived here for my entire life. The Johannesburg Stock Exchange doesn't use a comma as a decimal point, shops, banks, government institutions, etc -- NONE of them use a comma as a decimal point.

My South Africn bank statements (Absa bank) use the comma (if posted) and the full-stop (if downloaded). If you've never seen the decimal comma it simply means that you live in a publication bubble that uses only the full-stop. Many publications use the comma. The link most likely refers to "Supplement 1 of the Government Gazette R1146 of 5 July 1974". ABSA Bank (a large bank in South Africa) uses the decimal comma in their annual report []. -- 105.224.83.106 (talk) 08:03, 7 August 2014 (UTC) (leuce, not signed in)

Using a comma as a decimal point here would get you some extremely odd looks. — Preceding unsigned comment added by 41.135.110.211 (talk) 22:35, 20 August 2012 (UTC)

This: https://www.sadev.co.za/content/how-correctly-format-currency-south-africa has reference to the comma as the decimal point in South Africa. Avi8tor (talk) 16:13, 8 January 2023 (UTC)

The South African government editorial style guide at: https://www.gcis.gov.za/sites/default/files/docs/resourcecentre/guidelines/Editorial_Style_Guide.pdf states:

• Write decimal and negative numbers as numerals: 3,3 and -4. Use the decimal comma*, not the decimal point: 17,4

million.

• When South Africa adopted the metric system, it also adopted the comma as its decimal mark. However, be

alert to the fact that many other countries, use the decimal point.

• Use a space, not commas, to indicate thousands: 3 000; 20 000. Avi8tor (talk) 16:40, 8 January 2023 (UTC)

Postal number and ID-numbers are grouped in Sweden, like this:

• 703 75 instead of 70375 for postal codes, and
• YYMMDD-XXXX for personal numbers.

Is this a reason to change the text? Moberg (talk) 14:58, 12 April 2010 (UTC)

Decimal mark? I've never heard this term in all my engineering years. Moreover,

http://www.google.com/search?client=safari&rls=en&q=%22decimal+mark%22&ie=UTF-8&oe=UTF-8 shows <9000 instances of "decimal mark"

vs.

http://www.google.com/search?num=100&hl=en&client=safari&rls=en&q=%22decimal+point%22&aq=f&aqi=&aql=&oq=&gs_rfai= which shows > 1,420,000 instances of "decimal point"

This title of this page should be fixed, and decimal mark should redirect to decimal point, instead of the other way around. This is clearly the world-wide preferred phrase. —Preceding unsigned comment added by 94.252.120.217 (talk) 11:32, 9 December 2010 (UTC)

You call it the DECIMAL POINT because you use the POINT as the decimal MARK. Other cultures use other MARKS. HkCaGu (talk) 12:36, 9 December 2010 (UTC)

Nonetheless, it's name in English is "decimal point". It's confusing the issue to call it otherwise. I agree that the comma is the mark in other languages, but that's not a reason to present a page that misleadingly implies that the correct name is "decimal mark". Maybe there's a way to make this point clearer? Even the very first text on the page seems confused: "For the proper choice of the decimal mark in English language Wikipedia articles, see Wikipedia:Manual of Style#Decimal points." —Preceding unsigned comment added by 94.252.120.217 ( talk) 13:15, 9 December 2010 (UTC)

1. Have an empty article redirect from decimal point to decimal mark. 2. This is an article who's FOCUS is that typographical "thing" between the integer part and the fractional part. In some languages that "thing" is a full stop, in other languages something else. There must be no preferances from start in an encyclopedic article. How many of those > 1 420 000 instances have this focus? You can't use statistics like that. — Preceding unsigned comment added by 83.223.9.101 (talk) 12:25, 23 August 2012 (UTC)

The Bureau International des Poids et Mesures states: ...5.3.4 Formatting numbers, and the decimal marker The symbol used to separate the integral part of a number from its decimal part is called the decimal marker. Following the 22nd CGPM (2003, Resolution 10), the decimal marker “shall be either the point on the line or the comma on the line.” The decimal marker chosen should be that which is customary in the context concerned. If the number is between +1 and −1, then the decimal marker is always preceded by a zero. (−0.234, but not −.234) Following the 9th CGPM (1948, Resolution 7) and the 22nd CGPM (2003, Resolution 10), for numbers with many digits the digits may be divided into groups of three by a thin space, in order to facilitate reading. Neither dots nor commas are inserted in the spaces between groups of three. (43 279.168 29, but not 43,279.168,29) However, when there are only four digits before or after the decimal marker, it is customary not to use a space to isolate a single digit. (either 3279.1683 or 3 279.168 3) The practice of grouping digits in this way is a matter of choice; it is not always followed in certain specialized applications such as engineering drawings, financial statements, and scripts to be read by a computer. For numbers in a table, the format used should not vary within one column. — Preceding unsigned comment added by 83.223.9.101 (talk) 12:58, 27 August 2012 (UTC)

Similarly, the article states that the phrase "decimal dot" is a common informal term. Is this in the UK? As an American I have never heard anything other than "decimal point" in any context. Wschart (talk) 14:34, 23 November 2015 (UTC)

No, it's not common in the UK, though I suppose it might be explained as a "dot" to a young child learning how to write money. The correct term taught to all school children following the National Curriculum is "decimal point" Dbfirs 20:58, 23 November 2015 (UTC)

Obviously "decimal point" would be inappropriate as the title of this article, but I am wondering why it was not called "decimal sign" instead of "decimal mark"? — RFST (talk) 06:29, 2 January 2018 (UTC)

"Decimal sign" appears to be the older phrase, eg this book from 1862. By way of wikipedia article name popularity, the good ol' ASCII table lists only three marks: Exclamation mark !, Quotation mark ", Question mark ?, versus nine signs: Number sign #, Dollar sign $, Percent sign %, Plus sign +, Less-than sign <, Equals sign =, Greater-than sign >, At sign @, Not sign ¬. Batternut (talk) 22:43, 2 January 2018 (UTC)

That would seem to favour "sign" over "mark", except for a fan of switching things around just for the heck of it. My specific concern was "decimal point" for C/C++ locale information (probably named in the U.S.A.) vs. "decimal sign" in ISO 8601, until I found myself redirected to "decimal mark" in Wikipedia, something I don't recall seeing or hearing before. — RFST (talk) 11:38, 7 January 2018 (UTC)

"Decimal point" is always the term in the UK, so I'd never heard the term "decimal mark" before. Reconsidering the above list of marks and signs though, most of them relate to specific representations. By analogy to the generic article currency symbol though, perhaps "decimal symbol" would be the best generic term. Batternut (talk) 13:53, 7 January 2018 (UTC)

"Decimal symbol" does not seem appropriate because the glyph doesn't really represent something. Unicode seems like a good reference, and searching with Google for domain unicode.org yielded: 1 result for "decimal mark" (but only because the given name Mark occurred after "decimal"...), 2 results for "decimal marker", 3 results for "decimal comma" (all about using a comma character), a disappointing 4 results for "decimal sign", 6 results for "decimal symbol" (some duplicates there), 8 results for "decimal dot" (mostly "raised decimal dot"), a disappointing but understandable 196 results for "decimal point" (obvious English bias), and 387 results for "decimal separator" (including "DECIMAL SEPARATOR KEY SYMBOL", already mentioned in the article). Because of this I now amend my proposal to rename the article to "Decimal separator", with a brief section discussing the other names also used, and without choosing sides between the monosyllabic words "mark" and "sign" that people would probably prefer in everyday usage (here reluctantly listed alphabetically). — RFST (talk) 05:46, 9 January 2018 (UTC)

"Decimal separator" seems to be a good option, yes. Batternut (talk) 10:03, 9 January 2018 (UTC)

I made this a new section because renaming seems kind of important, and looked at some more statistics on iso.org (even though this ignores the protected texts of the standards themselves): 0 for "dot", 4 for "marker", 6 for "mark"/"symbol", 9 for "comma", 37 for "sign", 72 for "separator", 135 for "point". My interpretation is that, with "decimal point" disqualified for bias, "decimal mark" clearly has insufficient prominence to remain the title of this article (even less than "decimal comma"!), and that "decimal separator" again has higher adoption than "decimal sign". — RFST (talk) 04:16, 10 January 2018 (UTC)

This article could probably do with some additional info on the grouping of digits in the fractional part of a number. What is the prevalence of fraction grouping in various countries/cultures? It seems that, in most cases, the practice is to use no grouping. Are there places that commonly do? BigBadaboom0 (talk) 14:12, 27 January 2011 (UTC)

see International Bureau of Weights and Measures (Bureau International de Poids et Mesures) The International System of Units (SI) pdf section 5.3.4 "Formatting numbers, and the decimal marker," page 133 Zyxwv99 (talk) 13:51, 24 September 2012 (UTC)

NIST has this to say in its' guide to the SI-system: "10.5.2 Decimal sign or marker The recommended decimal sign or marker for use in the United States is the dot on the line [3, 6]. For numbers less than one, a zero is written before the decimal marker. For example, 0.25 s is the correct form, not .25 s.

10.5.3 Grouping digits Because the comma is widely used as the decimal marker outside the United States, it should not be used to separate digits into groups of three. Instead, digits should be separated into groups of three, counting from the decimal marker towards the left and right, by the use of a thin, fixed space. However, this practice is not usually followed for numbers having only four digits on either side of the decimal marker except when uniformity in a table is desired.

Examples: 76 483 522 but not:76,483,522 43 279.168 29 but not:43,279.168 29 8012 or 8 012 but not:8,012 0.491 722 3 is highly preferred to:0.4917223 0.5947 or 0.594 7 but not:0.59 47 8012.5947 or 8 012.594 7 but not:8 012.5947 or 8012.594 7

Note:The practice of using a space to group digits is not usually followed in certain specialized applications, such as engineering drawings and financial statements." — Preceding unsigned comment added by 83.223.9.101 (talk) 12:10, 19 October 2015 (UTC)

Could someone explain the bottom-left corner of this poster from 1905: http://upload.wikimedia.org/wikipedia/commons/a/a2/The_great_Victorina_Troupe.jpg I don't know hich country it's from (obviously an English-speaking one that used dollars in 1905), but they seem to use a period (.) to delimit the thousands and an underline for the decimals. This article does nothing to give me an indication as to where it is from.

(i.e, could someone edit the 'history' section to explain where and when this poster is from) — Preceding unsigned comment added by 132.165.76.2 (talk) 13:11, 22 December 2011 (UTC)

It's from USA. Printed by The Donaldson Litho Co. c. 1905, which you can find out in two minutes if you know how to use a search engine like, for instance, Google (how I miss AltaVista!). BTW, this is not a Q&A page. --Episcophagus (talk) 19:05, 5 November 2012 (UTC)

Similarly, in hexadecimal (base 16), full spaces are usually used to group digits into twos.

I've never seen groupings of two - always four. Windows Calculator uses four when digit grouping is enabled. Simple fix is to change "twos" to "twos or fours" but I'm curious if anybody actually uses two digits?

Poster Nutbag (talk) 14:37, 14 February 2012 (UTC)

As a programmer, I'd say grouping is twos (if grouping at all) is most common. This is probably because two hex digits corresponds to 1 byte (8 bits). You'd use fours if the natural way to group is pairs of bytes. This is not common since numbers (signed or unsigned) are more usually stored in more than two bytes in most modern programming languages (for example Java and the JVM have 4-byte 'integer' or 8-byte 'long'; there's also a 2-byte 'short' but this is less used; .NET has similar types.). Some file formats and wire protocols use numbers in the range 0-65535 (i.e. an unsigned two byte number, hence 4 hex digits). But when describing these you are more likely to group irregularly to show the meaning of various blocks of bytes (an example: https://en.wikipedia.org/wiki/PDB_%28Palm_OS%29). One of the more common exceptions, that would group 4 hex digits (2 bytes), would be representing Unicode (in particular UTF-16) as hex. 86.140.81.184 (talk) 09:55, 15 December 2015 (UTC)

This is undoubtedly a matter of calculator vis-à-vis computer convention for which provenance is needed. kencf0618 (talk) 17:27, 15 September 2022 (UTC)

What the heck is a "units digit" as mentioned in the history section? It's linkified to point at the digit page, which is of no help. Without a specific definition this passage doesn't make any sense. Monolith2 (talk) 16:24, 4 September 2012 (UTC)

The ones place, as opposed to the tens, hundreds, tenths, etc. — kwami (talk) 02:18, 5 September 2012 (UTC)

It's like the "unit block" referred to in police and fire department reports. If you live at 14 Maple Street, you don't live on the 200 block or the 100 block, you live on the unit block. Zyxwv99 (talk) 13:54, 24 September 2012 (UTC)

Consider the number 16,735.2809; the five is the units digit, the three is the tens digit, the seven is the hundreds digit, the two is the tenths digit, etc. P.S. talk of 100 block, 200 block, etc. won't make sense to those unfamiliar with North American addresses. In Australia, for example, the lots are numbered consecutively (without skipping numbers) ... it's less convenient. JIMp _talk·cont 01:28, 10 October 2012 (UTC)

This article says: "Gerbert of Aurillac marked triples of columns with an arc (called a "Pythagorean arc")". However, according to History of Mathematics, Volume 2, p. 177 by David E. Smith a "Pythagorean arc" is the same as the abacus itself. --Episcophagus (talk) 18:53, 5 November 2012 (UTC)

There seems to be consensus about separating thousands with a small space, but what does it mean precisely in terms of Unicode character? I haven’t found official recommendation about this.

• The page Space (punctuation) says that the Unicode Thin Space is “[r]ecommended for use as a thousands separator for measures made with SI units. Unlike U+2002 to U+2008, its width may get adjusted in typesetting.[24]”, with the ref pointing out to “The Unicode Standard 5.0, printed edition, p.205”. It is available here, if I am not mistaken. This document does not mention thousands separators, thus the reference only covers the “adjustable width” part of the quote here above. This leaves the first claim, that Unicode Thin Space is recommended for use as a thousands separator, without source. I do not think that the Unicode Thin Space would be recommended for use as a thousands separator by anybody (or by any serious body), because it is breakable.
• If one insists on separating thousands with Unicode Thin Space, then it should probably be complemented with a Unicode Word joiner (“The word joiner can be used to prevent line breaking with other characters that do not have nonbreaking variants, such as U+2009 thin space or U+2015 horizontal bar, by bracketing the character.”, section 16.2, p. 534, Unicode Standard 5.0). But this seems far-fetched, and just my idea, there is no evidence (that I found) that it is recommended by any official body.
• According to the wikipedia page about ISO 31-0, “small space” is recommended for thousands separation. It may be that ISO 80000-1:2009 talks about this (because it supersedes ISO 31-0), but I do not have access to it. Can someone check in there if anything more precise is specified?
• The International Bureau of Weights and Measures (referenced in Decimal_mark#Digit_grouping) indicates the following. “Following the 9th CGPM (1948, Resolution 7) and the 22nd CGPM (2003, Resolution 10), for numbers with many digits the digits may be divided into groups of three by a thin space, in order to facilitate reading. Neither dots nor commas are inserted in the spaces between groups of three.” There is no indication that the phrase “thin space” there refers to a Unicode Thin Space, it should probably be understood non technically, as a space that is not large but thin. Also, Resolutions 7 and 10 do not use the wording “thin space”.

It would be good to find some official recommendation about which character to use as a thousands separator. I can only think of Unicode Narrow No-Break Space as a proper separator for thousands. But maybe there are other opinions out there.--OlivierMiR (talk) 12:51, 13 May 2013 (UTC)

Note that this question is related to, but different than, the question of separating numbers and units, see Talk:Space_(punctuation)#Reference_for_thin_space_between_numerals_and_units.--OlivierMiR (talk) 13:04, 13 May 2013 (UTC)

I concur. I agree that although there is consensus about using a small/narrow/thin space, there seems to be little or no instruction on which Unicode code point to use for it. But I feel that it makes sense and is only natural that it happened this way, at least so far, because the glyph-vs-character distinction is at work here. The cited standards are actually all talking about the glyph that they want, not about character encoding, although this fact is implicit rather than explicit (and many of their authors would not have known how to say it in those words). All that the content people (authors, readers) truly care about is the end result, namely, the glyph appearance, the line-break logic, and how much it costs to get it (as opposed to what mechanism accomplishes those); the question of what digital typography to use under the hood to achieve that end result, namely, which character encoding to use, is up to the person responsible for the encoding, whether it be a typesetter/compositor, desktop publisher, webmaster, or other "technical person". This only makes sense when you think about the fact that typesetting has changed so much over the past 4 decades. In 1974 they would have been talking about the best way to set it in hot metal on a Linotype machine; in 1984 they would have been talking about how best to achieve it in photocomposition on film; in 1994 they would have been griping about the limitations of ASCII and ISO/IEC 8859; only today do we have the luxury of bringing Unicode into the discussion and expecting it to apply widely across print and web. So it only makes sense that, for example, the AMA Manual of Style says "thin space" but doesn't mention encoding or code points. It was written by non-typesetters who expected the typesetting to be left to "the technical people", which was just as well, because the means to the end kept changing repeatedly with evolving technology. But certainly it seems plausible that maybe we've arrived now at a mature architecture that will be around for decades (Unicode encoding of digital documents), so maybe future editions of major style guides will mention particular Unicode points that they want to convene upon for particular semantic uses. I agree with you that Unicode Narrow No-Break Space (U+202F) seems like the better option for a digit-grouping character (thousands separator) than the Unicode Thin Space (U+2009) because it looks similar but is nonbreaking. But it would take someone with greater character encoding and web development expertise than I to know for sure whether there is any reason to avoid using either of these characters. One last thought that is interesting and relevant. In pre-Web professional publishing, back when people were willing to pay for human typesetters to crawl over the column of text and tweak the line breaks into aesthetic nicety, it wouldn't have mattered very much whether the breaking or nonbreaking character was used. But of course people realized that rightfully, that logic should and could be automated, and eventually there comes the era when "no business firm can afford to pay for the non-automated method anymore." Quercus solaris (talk) 22:24, 30 July 2013 (UTC)

What about the American (?) custom of writing "0.5" as ".5", that is "a symbol used to separate the integer part nothing from the fractional part of a number"? It looks quite weird for people from Europe, so maybe somebody can write a referenced summary about such use, its history and what various standards/style manuals think about it? — Mikhail Ryazanov (talk) 00:31, 22 June 2013 (UTC)

The omission of the leading zero from decimal numbers less than 1.0 has several contexts of appearance. The widest context is in unedited usage in which people (almost anyone, not just Americans) are simply not being fastidious about typing the leading zero. But another context of usage, for which there are plenty of references available, occurs also in edited and published usage, which is in statistics when talking about quantities that can never be more than 1.0, such as α, β, and P. Various house styles drop the leading zeros on such values, including AMA style and APA style. I don't agree that either of the above contexts (the carefree unedited one or the edited statistical one) are unfamiliar to Europeans in any generalized, stereotyped way. As for AMA and APA style being American, well, there is a global STM audience that encounters them on a daily basis. They may simply be unfamiliar to some individual Europeans. Quercus solaris (talk) 02:09, 29 July 2013 (UTC)

This article talks about countries instead of languages. Even though in many countries there might be an "official" decimal mark (and this will be given by law), in many cases the authorities of the language (and not a government) determine what should be used. Then, it makes no sense to say that in Spain the style "1.234.567,89" is older. According to the reference (RAE), this happens in Spanish (everywhere in the world), and this does not necessarily happen in Spain in any of the other official languages. Another example: French/English Canada. In many cases, the writing depends on the language, not the country (should I write 1˙234˙567,89 if I write in English but I am in Italy?). — Preceding unsigned comment added by 88.17.128.16 (talk) 13:33, 16 March 2014 (UTC)

"In many cases, the writing depends on the language, not the country" - but in many cases, the writing depends on the country in addition to the language (i.e., the language alone is not sufficient to know the decimal mark style). So it is useful to group by country, and for multi-lingual countries to then fork by language. The reason why the country is important (Swiss, Austrian and German decimal notation for German texts differ significantly!) is that the main "trend-setters" for notation - namely norms and school books - are usually local to countries, and not languages.

However, there is certainly no structure for this article that is simple and satisfying for all cases.

--haraldmmueller 15:23, 10 November 2014 (UTC) — Preceding unsigned comment added by Haraldmmueller (talk • contribs)

This section of the article claims that in countries (basically the Chinese culturosphere) which have numerals which go up in 10000s, the separator divides groups of four digits. Although this would be obviously the rational way to do it, at least in Japan this claim is totally false. It is essentially never done (with the possible exception of rational foreigners); the commas are put every three digits, and people perform these amazing power conversions in their heads, to read for example 100,000,000 as 'ichi-oku' (i.e. one ... 10^8) You will even see large amounts of money written in units of, say 10万 (10...10^4 : i.e. 10^5), and the numbers will be written as 12,340, (i.e. 12 340 00000 = 1 234 000 000), leaving the comma before 10^(3+5)). If someone can confirm practice in Korean or Chinese, we can probably replace this. Imaginatorium (talk) 10:03, 3 July 2014 (UTC)

Just a followup: Chinese WP, via Google translate because my knowledge of Chinese is very restricted (though nonzero), in Chinese the separation by four digits is *sometimes* done. Imaginatorium (talk) 10:12, 3 July 2014 (UTC)

• I modified that part. As far as I know 4-digit grouping is not that common in China. To convince myself I just browsed through a several Chinese wikipedia pages that contain large numbers and digits are grouped by 3. But we could definitely use an input on that. Now I know for a fact that 4-digit grouping is never done in South Korea (and presumably North) and although there have been suggestions that 4-digit grouping would make sense, that is only speaking in theory. Now since I cannot find a suitable source for East Asian countries maybe I should remove it, but I thought the info that those languages count by myriads is useful and since this page has been spreading misinformation for some time I just changed the sentence to the opposite. If people can find a good citation they can remove that tag, alternatively if they choose to remove the whole sentence I am OK with that. PS: on the kanji separator comment above, similar stuff happens in Korea as well, but I do not see them as grouping characters. It is just another way of writing: the equivalent of "thirty-seven million and five hundred twelve thousand fifty" as "37 million 512 thousand 50" it does not happen in English, but it does in those languages. IdentityCrisis (talk) 20:53, 20 October 2014 (UTC)

• In Chinese, 1234567890 is read as 12亿3456万7890, so four-digit grouping like 12 3456 7890 is theoretically the most suitable for Chinese. However, the Chinese national standard GB/T 15835-2011 "General rules for writing numerals in public texts", item 5.1.1 states that three-digit grouping should be used for large numbers, separated by either commas or spaces. They are respectively called 千分撇 (thousands-separating comma) and 千分空 (thousands-separating space). --Yejianfei (talk) 01:30, 11 December 2024 (UTC)

Someone added into the text: "The sorce (sic) [19] is a book from 1923 - so reliable to (sic) the internet age? Never seen in real life that one uses 1.234,567.890,12."

(a) There are discussions about texts from 1905 above. 1923 is well within the time where lots of texts survive up to now - please finde the respective fields yourself.

(b) The book's reliability does not change with the "internet age". The number is written as it is.

(c) I have learned this way of digit separation in school in the 1970s. However, this is not a valid citation for WP, so I looked for some real use. I did not do an extensive scientific search for newer documentation, e.g. school books - but take it for granted that it was used like this much longer than 1923. If you look into the article, you will see many more claims that are not at all supported by any evidence.

(d) "Never seen in real life that one uses ...". I am happy to learn about all the German and Austrian texts you have read, and the authors of books with such numbers you had personal contact with to claim that your "(I've) never seen ... that one uses ..." is a relevant portion of German and Austrian writing culture. Please cite the research you did that shows that my addition (of a factual use of signs) is, at some relevant time, never used.

Cheers.

--haraldmmueller 19:54, 9 September 2014 (UTC) — Preceding unsigned comment added by Haraldmmueller (talk • contribs)

I grew up in England (1950s-60s) and knew only the raised dot for decimals. As I recall, bank statements were computer printed using a raised dot (which in those day meant a systems programming operation!) into the 1970s or 80s. My guess is that this is the traditional system for the non-US English speaking world, but not anywhere else. So I have removed Japan from the list, and someone should check Korea, which sounds suspicious. The only other consideration is that I believe it is possible in vertical writing, to write decimal numbers using kanji, and then the decimal mark would be a central dot (but this is not the same as the raised decimal point). Perhaps vertical writing could/should be added as images, as I doubt if en:WP has the facility for vertical text. Imaginatorium (talk) 08:54, 3 November 2014 (UTC)

Yes, the Interpunct (centred dot) was standard in the UK until 1968 except on typewriters. Dbfirs 21:09, 23 November 2015 (UTC)

This is a misreading of the cited reference from 1968. It says that a decision was made between the decimal point and the decimal comma, and the decimal point was chosen. At the time everybody would have understood the decimal point to mean the middle dot. — Preceding unsigned comment added by 91.110.51.189 (talk) 12:25, 21 May 2023 (UTC)

These appear to be using groupings of 4 decimal digits, with no obvious special significance to the groups. — Preceding unsigned comment added by 38.122.49.58 (talk) 17:32, 10 November 2014 (UTC)

Yes, but these are not representations of numbers (in the mathematical sense); they are simply strings of numerals. I do not think even mention of this is relevant at all to "decimal point (mark)". Imaginatorium (talk) 03:13, 11 November 2014 (UTC)

I am spanish, and I can give some information about that "(In Spain, in handwriting it is also common to use an upper comma: 1.234.567'89)[citation needed]" I prefer to share it here instead of editing the article.

I am Physics and Chemistry teacher, and sometimes I have teached mathematics, and despite I don't do it, I can confirm a lot of mathematics teachers do it in blackboard, they say it is more clear to write/explain range between numbers. For example they prefer (4'5 , 6,7) instead of (4,5 ; 6,7)

The RAE (Real Academia Española = Spanish (Language) Royal Academy) says it is not correct, so it somehow admits it happens. Article about "comma"

http://lema.rae.es/dpd/srv/search?id=V1EqcYbX4D61AWBBrd#4

"4. Usos no lingüísticos En las expresiones numéricas escritas con cifras, la normativa internacional establece el uso de la coma para separar la parte entera de la parte decimal. La coma debe escribirse en la parte inferior del renglón, nunca en la parte superior: π = 3,1416. Pero también se acepta el uso anglosajón del punto, normal en algunos países hispanoamericanos (→ punto, 4.4): π = 3.1416."

English translation: "Non linguistic uses: In the numerical expressions written in figures, international regulations require use of the comma to separate the integer part of the decimal part. The comma should be written at the bottom of the line, *never on top*: π = 3.1416. But the Anglo-Saxon point using normal in some Latin American countries (→ point 4.4) is also accepted. Π = 3.1416 "

It is something common in handwriting (blackboard), not very common in books / documents, so it is hard to give a reference. Just as some kind of official citation, I can give some official documents examples where upper comma is also used: chemistry exams for university access http://www.juntadeandalucia.es/innovacioncienciayempresa/sguit/paginas/distrito/examenes_sel_m25/selectividad_2011/sel_2011_quimica.zip

Enrique.garciasimon (talk) 00:29, 12 January 2015 (UTC)

Most people would simply like to know the correct modern notation to use. This is not summarised in the intro, not is it ever clearly spelled out in the article. I was trying to discover the correct British notation but after reading the lengthy paragraph on this subject the article failed to ever tell me what the correct notation actually is.

Regarding the decimal mark, the answer on British style is already there. It's at Countries using Arabic numerals with decimal point > [Alpha list] > United Kingdom. Regarding the word "correct", all of the options discussed (for both decimal mark and digit grouping) are correct; which one is normative depends on language, region, and (for digit grouping) style guide. Regarding digit grouping, the article could use more clarity because the topic itself could use the same. It is not standardized by either language=English or country=UK. There are only prevailing styles to choose from, and you have to decide whose style you'll follow. For the UK, the prevailing style for digit grouping (according to both New Hart's Rules and Butcher's Copy-editing, which are two of the leading UK styles) is divided between general/nontechnical copy and scientific/technical copy. Neither of those guides says "such-and-such is correct or incorrect." They both just state a variety of things that are usually done. They both say that general copy usually uses commas and technical copy usually uses space or thin space. As for 4-or-more versus 5-or-more, New Hart's says that general copy usually uses commas for 4 digits or more (thus 1,000 and 10,000) and that technical copy uses thin space (thus 1 000 and 10 000). Butcher's says that general copy usually uses commas with either a 4-or-more scheme or a 5-or-more scheme (take your pick) but that technical copy usually uses a space and a 5-or-more scheme (thus 1000 but 10 000). Both guides state that whatever style is used in text, it should be overridden as needed in tabular matter to enforce decimal-place alignment (up and down the column). By now you may be wondering why the hell there isn't one simple rule instead of all that variability just described. The answer is that no single authority governs it in English, either UK English or US English. Only prevailing orthographic consensus(es) and house styles govern it. Welcome to the human condition—negotiations, power struggles, people who think that X is "wrong" when it's really not, people who won't agree to standardize on Y partly because they like Z better but also (although they won't admit it to others and possibly not even to themselves) rather more because they want to keep the guy who likes Y from being seen having his way and they want to claim that their way is better so that they can be seen having it and get to enjoy making guy Y bow to it. See also a century of wasting resources on anti-metrication "logic". Quercus solaris (talk) 22:41, 17 April 2015 (UTC)

The SI style (with thin spaces) is becoming more common in the UK, and is usually taught in schools, but people like to cling to the older thousands separator. Dbfirs 16:38, 10 May 2015 (UTC)

Some NLS systems allow to specify different decimal separators for different contexts even within the same locale:

• Numeric decimal separator - for "normal" numbers, typically a comma or dot, as already discussed in the article.
• Monetary decimal separator - for usage of numbers in conjunction with currencies, sometimes the same as the numeric decimal separator, sometimes the opposite of either comma or dot, sometimes another character such as the corresponding currency symbol. Example: In Portugual, the comma is the normal decimal separator, but the cifrão sign ($) was used in currencies instead of the decimal comma, as in 2$50.
• Decimal version separator - the decimal separator used in version numbers (if interpreted as decimal numbers), almost always a dot (even if the numeric decimal separator is a comma), but I've also seen spaces, hyphens and slashes. Example: In Germany, the comma is used as the normal decimal separator, but a dot is used for version numbers. (I've also seen this being implemented to automatically use either the same character or the "opposite" as the numeric decimal separator, dot or comma, but so far I have considered this to be a bug.)

It might be interesting for readers to find some further background on this in the article. Which conventions exist globally? Are there any standards for this? Are there even more such "contexts" where decimal separators, differing from the default ones in the corresponding locales, are used? --Matthiaspaul (talk) 01:42, 15 December 2016 (UTC)

The map is wrong for Spain. From 2010 the Real Academia de la Lengua recommends the usage of the decimal point. The CSIC style guide recommends it from much earlier. — Preceding unsigned comment added by 90.71.73.34 (talk) 22:41, 15 May 2017 (UTC)

The page for Set notation doesn't discuss why, but it uses commas to separate {π, 6, 1/2}. I'd assume there would have to be a difference for those that use a comma as this could also be written as {π, 6, 0.5} or {π, 6, 0,5}, in which case the comma becomes ambigious. — Preceding unsigned comment added by 80.193.174.53 (talk) 15:52, 24 May 2017 (UTC)

I think in general that is disambiguated by placing a space after the comma when enumerating the set. But I looked at a section of the French article, fr:ensemble#Ensembles finis ("Finite sets"), and there are a couple of uses that are without spacingh:

• la paire des éléments 1 et 2, notée {1,2}, ne doit pas être confondue avec le [[couple (mathématiques)|couple]] (1,2), qui a un ordre déterminé.
• The [unordered] pair of elements 1 and 2, written {1,2}, is not to be confused with the ordered pair (1,2).

This may just be a typo in French Wikipédia, though. The article gives no examples with decimal fractions. I checked German and Hungarian but the examples of set notation there are either with natural numbers, or with algebraic letters. Perhaps sets of real numbers are not commonly enumerated so the point is moot? 94.21.219.120 (talk) 15:47, 9 December 2018 (UTC)

On the website of the Namibian Government, they use the dot as decimal seperator: http://www.gov.na/population Makes sense to me, since they are an English speaking country. So I would ask to move Namibia from "countries ... with decimal comma" to "countries ... with decimal point" — Preceding unsigned comment added by 2003:c0:e3de:f775:39b7:19a1:cddc:d14 (talk) 00:15, 27 January 2018‎ (UTC)

Agreed, and changed. Batternut (talk) 23:19, 27 January 2018 (UTC)

Both comma and point are used, within gov.na and outside it. Status quo is preferable until better evidence is found. Or perhaps there should be a third category. — RFST (talk) 06:28, 28 January 2018 (UTC)

Yes, South Africa would also fall into a mixed category, judging by the comments at Wikipedia talk:WikiProject South Africa#Should South Africa articles use "continental system" numbers?. Perhaps a section entitled something like "Countries using Arabic numerals with both decimal comma and decimal point". Batternut (talk) 13:53, 28 January 2018 (UTC)

Perhaps, but that's rather a long title, and it seems to be a declaration that you have full freedom (without even so much as a style guide, see talk section about South Africa), so it should at least be worded differently. And how much divergent use is enough to put a country in this category and on a slippery slope? And is it really only by country, or also by language (a computer locale is first by language and only then by country, and maybe some information should be sourced from computer-related projects)? — RFST (talk) 06:39, 29 January 2018 (UTC)

You raise three good points. The section name is indeed terribly long. Since both current sections are in the "Hindu–Arabic numeral system" section, one option could be to strip out the implied "..using Arabic numerals...", giving subsections "Countries using the decimal point" ,"Countries using the decimal point" and "Countries using both decimal comma and decimal point".

Or perhaps another approach entirely - to replace both sections with a single table, one row per country, with a column list of "country", "decimal point usage" and "decimal comma usage". Ticks and crosses or some text equivalents would generally suffice; for countries using both though any kind of qualification could be added. For the South African example, in the comma column, we might mention the official position, while in the point column observing its predominance. Complicated cases could use section links to expanded descriptions lower in the article. Per language variation could also be mentioned in the table, and expanded upon below. Batternut (talk) 09:15, 29 January 2018 (UTC)

A single table probably takes too much vertical space and is still not very informative compared to a structure that would show the putative causes (luckily there's a picture, although it has Namibia in blue). There are already several countries mentioned in both sections (Canada, Luxembourg, Switzerland, perhaps more) with proper qualification; you could perhaps do the same with Namibia and South Africa. — RFST (talk) 15:21, 29 January 2018 (UTC)

That would probably be fine, potentially with footnotes. Batternut (talk) 16:30, 29 January 2018 (UTC)

Now in both lists, with qualification "(uses both marks)" and a reference for the decimal comma. Perhaps it would be useful to describe usage further, because references with decimal comma are far more easily found than for South Africa but perhaps also in a decline. — RFST (talk) 06:44, 4 February 2018 (UTC)

"ALGOL ended up allowing different decimal separators"

Can this be true? It seems it would wreak havoc parsing argument lists to have commas in literals. (More likely they permitted different formats in i/o, not in the actual source code. ) Gjxj (talk) 14:22, 22 March 2018 (UTC)

I am confused. Why is "Digit grouping" a section under the article "Decimal separator"? The topic is about the separation of the integral and fractional parts of the number, not the grouping of the integer. —Wei4Green | 唯绿远大 (talk) 20:35, 10 November 2018 (UTC)

I was wondering the same thing. (There are several redirects to that section, too.)

The two are closely related, the symbol used for digit grouping must needs be different from that used as a decimal separator, so splitting the article may do more harm than good.

Treatment of the fractional part may also require grouping symbols. Strict SI orthography does, using a thin space as the group separator, but I have rarely seen it used in English-language texts, and this is not mentioned at the SI article. (See https://physics.nist.gov/cuu/Units/checklist.html item #16, "The digits of numerical values having more than four digits on either side of the decimal marker are separated into groups of three using a thin, fixed space counting from both the left and right of the decimal marker.") For quantities expressed in SI, it should be far more likely to have separators in the fractional part, since the integer part should be be expressed in terms of 10³ⁿ such that the integer part is never more than three digits, anyway.

It might be worth formally proposing a split. 94.21.219.120 (talk) 15:21, 9 December 2018 (UTC)

I, too, was wondering why the topic doesn't have its own page. Note that it gets its own page in some other languages' Wikipedias (e.g. in German), which could be drawn upon. Regarding the concern above that "two are closely related, the symbol used for digit grouping must needs be different from that used as a decimal separator, so splitting the article may do more harm than good": One option would be to leave a summarizing section that focuses largely on that relationship, but put a full article link at the top of the section. —Undomelin (talk) 22:25, 3 May 2022 (UTC)

Would the table on my talk page be a helpful addition?Lee De Cola (talk) 17:52, 7 January 2024 (UTC)

Austro-Hungarian Empire and Austria prior to 1938 used · Dieſelmaus (talk) 01:55, 15 November 2019 (UTC)

The radix point article starts like it is going to be more general than decimal separator, but it has practically no information about radix separators for other number bases, other than they may be used. Thus the information contained in the head section is basically a small subset of that contained in decimal separator.
On the other hand, the information in the body of the article is not about the radix character (which is how "radix point" is defined in the head) but rather on the representation of fractions in positional number notation. Therefore, that information should be merged into Positional_notation#Terminating_fractions, if it is not all there already.
--Jorge Stolfi (talk) 20:32, 4 July 2021 (UTC)

The following Wikimedia Commons file used on this page or its Wikidata item has been nominated for deletion:

• DecimalSeparator.svg

Participate in the deletion discussion at the nomination page. —Community Tech bot (talk) 14:22, 13 May 2022 (UTC)

It appears Australia has adoped the comma on the line and space for thousands. https://extranet.education.unimelb.edu.au/SME/TNMY/Decimals/Decimals/backinfo/overseas.htm Avi8tor (talk) 14:37, 8 January 2023 (UTC)

I haven't seen anyone using "." for grouping thousands in Denmark since I went to school, where it was taught. I doubt anyone uses it in real life, except when writing bills and such. 5.186.78.167 (talk) 11:32, 12 October 2023 (UTC)

Spanish Real Academia Española (RAE) recommend using a thin space as thousands separator, and a dot as decimal separator.

Usage of comma as decimal separator is discouraged, but also accepted, since using a space as thousand separator disambiguates its meaning. A dot is preferred to align with international standards.

However, this official rule is often not applied. Most people don't know about it or don't like it. Also it is common to find contradicting information (like this Wikipedia article, at the time of writing).

I think it would be nice to at least note in the article that there are conflicting versions and point to the official vs. popular formats.

Sources:

• Diccionario panhispánico de dudas: números
• Ortografía de la lengua española
• 20 minutos: Los números no se escriben como crees: la RAE explica cómo se hace y Arturo Pérez-Reverte da su versión

Gimix3 (talk) 11:02, 3 November 2023 (UTC)

That seems reasonable, and much of the above text could be used as-is, other than Wikipedia doesn't use contraction like "don't", the entire comment about "Most people" isn't encyclopedically relevant, and the citations would need to be done inline.  — SMcCandlish ☏ ¢ 😼  06:54, 4 November 2023 (UTC)

I can assure as mathematics high school teacher that a high percentage of the population uses the raised comma (0'23). 2A02:9130:1A9:8A74:7DE4:4472:2BDA:DC33 (talk) 06:20, 17 December 2024 (UTC)

It is said that there is no separator used in writing years. But the usage is more complicated. Yes, when writing out the year in many dating systems AD, CE, AH, BP etc.for dates of historical interest, which happen to be confined to four digits or less, no separator is used. Thus 2024 CE or 2234 BCE. But when - rarely - there is a reference to a year using more than four digits, such as 20,000 BP, then a separator is often used. On a different topic, when a number is using a numeral in some systems, such as month. day, year, or a base 60 system (eg Babylonian or degrees minutes seconds) a separator is needed. It gets more complicated than this article touches on. TomS TDotO (talk) 04:47, 13 February 2024 (UTC)

The V language can use the underscore (_) as a separator. However, it was summarily blocked from inclusion by certain opposing editors. Arguably unfair special new criteria or language preferences are imposed, beyond having a Wikipedia page and notability.

V, based on TIOBE and GitHub rankings, is verifiably about or more popular than Rebol, Red, D, OCaml, and Zig, which were languages placed in the article unchallenged and without abrasive reversions. Wukuendo (talk) 08:04, 28 May 2025 (UTC)

Central discussion at Wikipedia_talk:WikiProject_Computing#Significance_of_V_(Vlang)_and_Pony. Kindly keep it there rather than copy/pasting comments all over the place. MrOllie (talk) 13:32, 28 May 2025 (UTC)