Wikipedia talk:Manual of Style/Dates and numbers/Archive 87

Archive 80

←

Archive 85

Archive 86

Archive 87

Archive 88

Archive 89

Archive 90

"U.S." vs. "US" in the text of this guideline

See also Wikipedia talk:Manual of Style/Archive 81#U.S. & Wikipedia talk:Manual of Style#U.S. vs US.

Apparently, at least one person monitoring this project page thinks there is not a style guideline of this minor issue, but there is such guidance. The Wikipedia:Manual of Style (abbreviations) clearly states that the term "United States" should be abbreviated as "U.S.". First, I think this is the correct style; and second, I strongly suggest that the issue be settled in the MOS:ABB lest we see this be a repeat problem with other Wikipedians just trying to follow the MOS. --Charles Gaudette 07:19, 27 September 2007 (UTC)

First, this abbreviations submanual merely gives the ungainly and old-fashioned "you dot es dot" merely in a table square as an abbreviation, without further wording or proscription. The table was wrong or questionable in at least two other respects I noticed on a quick scan through: in failing to provide undotted "am" and "pm" as options—see MOSNUM and MOS on that (I've added those now); and in showing the undotted form "PhD", first, no less, against the explicit preference of MOS.

Second, such a proscription of "US", if it really appeared in explicit wording, would be in conflict with MOS itself, which says the following:

• Periods and spaces
• Many periods and spaces that were traditionally required have now dropped out of usage. For example, PhD is preferred to Ph.D. and Ph. D. Periods are retained in abbreviations that cannot otherwise be clearly identified.

Third, "US" is all over Wikipedia, possibly because people think it looks silly to have "the U.S. and the UK" together, and because just about every American initialism, including, hold your breath, "USA", has dropped the dots. Indeed, until some dot zealot tampered with it about a month ago, we had "US-registered" at the bottom of every WP page; now, it's "U.S. registered", and it now goes against correct practice, in the US and elsewhere, of hyphenating such an item, since the dots make it look pretty awful "U.S.-registered".

Fourth, our MOS is not Chicago's poodle, contrary to the clear implication right at the top of MOS:ABB. In any case, Chicago—excellent and authoritative as it (mostly) is—should get over it and change this visually unattractive and inconsistent practice: almost all other English speakers, and many US writers, have have moved with the times (excepting upper-case text, of course, where it may not be clearly identifiable—see MOS wording above). TONY (talk) 10:06, 27 September 2007 (UTC)

Fifth, there's nothing stopping you from using the dotted version if you want; but mass changes to existing articles for the sake of it may well be considered undesirable, as for other linguistic options. TONY (talk) 10:06, 27 September 2007 (UTC)

And PS, now you've changed it to "U.S. dollars (US$123)" in MOSNUM's Currencies section. <gives quizzical look> TONY (talk) 10:10, 27 September 2007 (UTC)

Sounds like a to-do item then: Make WP:MOSABB agree with WP:MOS. — SMcCandlish [talk] [cont] ‹(-¿-)› 16:37, 27 September 2007 (UTC)

Just a note here on Tony1's PS, the only reason I have "now" changed "U.S. dollars (US$123)" in MOSNUM's Currencies section is because I had not seen Tony1's comments and revert at that point in time. --Charles Gaudette 15:42, 27 September 2007 (UTC)

It is "U.S." – if you also write "N.A.S.A.", "U.K.", "S.C.U.B.A.", "I.B.M.", "N.Y.S.E.", etc. Which hardly anyone does any more. — SMcCandlish [talk] [cont] ‹(-¿-)› 14:34, 27 September 2007 (UTC)

... and L.A.S.E.R.? Thunderbird2 14:57, 27 September 2007 (UTC)

Sigh. Who edited my heading? Bad people, very bad. --Charles Gaudette 15:27, 27 September 2007 (UTC)

I did, if you mean addition of the parenthetical. Per WP:REFACTOR; no one owns headings, and they need to make sense in the context and help keep things on-topic per WP:TALK; I figured that would be obvious (and the rationale was given very clearly in the edit summary: This is MOSNUM after all, so the topic of "US" vs. "U.S." is only relevant here inasmuch as it relates to dates and numbers. :-) — SMcCandlish [talk] [cont] ‹(-¿-)› 15:39, 27 September 2007 (UTC)

The change to the heading does change the meaning of my comment. You are right, of course, I don't "own" anything here, but I do try and say-what-I-mean and mean-what-I-say. I was referring to the prose of the project page as it was reverted by Tony1. My mistake for not being clearer. --Charles Gaudette 15:51, 27 September 2007 (UTC)

Ah, I see. I misunderstood. Is it better now as ' "U.S." vs. "US" in the text of this guideline'? My goal was simply to prevent incoming editors from mistaking this topic as general discussion of what WP's style guide should recommended with regard to periods and abbreviations, which is already being discussed in too many places instead of centralized and settled (I would suggest that WT:MOS is the place for that, since WP:MOS is the controlling document and its subguidelines simply explorations of subtopics in more detail). — SMcCandlish [talk] [cont] ‹(-¿-)› 16:30, 27 September 2007 (UTC)

Q: "Is it better now as ' "U.S." vs. "US" in the text of this guideline'? A:Yes. ;-) Further, I am in agreement with your rational, SMcCandlish. I have (re-)read some of those old back-and-forths about "U.S." versus "US" and I don't want to repeat them, (I don't have that sort of free-time). --Charles Gaudette 18:38, 27 September 2007 (UTC)

Drat, there goes the opportunity for yet another perfectly good (re)edit war! ;-) Askari Mark (Talk) 19:09, 27 September 2007 (UTC)

I'm very happy with the dot-it-or-don't policy that has been around for a while. I'll buy every Wikipedian a beer if Americans at large haven't dropped the dots, overwhelmingly, by 2015. Tony (talk) 10:35, 30 September 2007 (UTC)

We haven't done so. That may well be true for USA as pointed out above (but while the U.S. Army omits it when USA means United States Army, the generally public probably doesn't use that very often). Longer acronyms are move easily recognised as such. And we generally have from measurement symbols such as US$ and quite often from US gal for the U.S. gallon, at least for those of us who cringe when we see "lb." and even worse "lbs.", but that doesn't affect the general rule. Gene Nygaard 15:25, 9 October 2007 (UTC)

Further to the above, the policy to use "U.S." has been around since at least 2004, with a short break recently when it was commented out, apparently by mistake, restored by me on the 6 September, removed by Tony on 7 September (who thought I was making a unilateral change (despite my edit summary <sigh>)) and restored by me today. Rich Farmbrough, 10:36 5 October 2007 (GMT).

(Outdent) Needs discussion. I've posted on Richard's talk page some reasons that reinstating these rules is of questionable value. Tony (talk) 11:12, 5 October 2007 (UTC)

For information, a compromise guideline now appears in MOS, specifying that the dotted version is more common in AmEng than the undotted, and that in the same variety of English, many authors avoid the abbreviation when the United States appears with other countries in the same sentence. (The second point is a Chicago formality that is widely not observed by Americans themselves). The list of abbreviations at WP:ABB has been updated (by Crissov, I think), to include "US" as an option alonside "U.S.". Tony (talk) 08:36, 8 October 2007 (UTC)

The 29th September

Why does the manual of style say not to use "the"? - Articles written in British english read funny without the "the"... --Fredrick day 23:34, 28 September 2007 (UTC)

Articles written in American English would read funny with the "the" (unless "of" is also used). Using the "th" (and similar) is discouraged by MOSNUM in favor of using auto-formated dates. — Aluvus t/c 05:30, 29 September 2007 (UTC)

"The" looks very funny in any variety of English. Tony (talk) 06:15, 29 September 2007 (UTC)

Agreed. And I would add that this is just one of those things where the formatting and how one says it aloud do not necessarily have a 1:1 relationship. Read aloud "He died 29 September 2007", will be said differently by different people (unless they are making a conscious effort to read precisely what is written before them). Some will say it as written, others will say "29th September", others will say "the 29th of September", etc. Similarly "US$57" is usually read aloud as "fifty-seven U S dollars", despite the way it is formatted, and most (though not all) speakers would audibly render "22 cm" as "twenty-two centimeters", not "twenty-two C M", though the latter version would be permissible and understandable to most. — SMcCandlish [talk] [cont] ‹(-¿-)› 22:56, 1 October 2007 (UTC)

Spacing after the decimal point

Who wants to see 0.249 592 9? Mark Askari has pointed out, when I queried it here before, that it's acceptable in some circles. But on WP I think it's unkind to the readers. I edit science, and until this came up I was unfamiliar with the practice of such spacing. I agree with SMcCandlish's change. Tony (talk) 15:02, 1 October 2007 (UTC)

I do. It is not just acceptable, but mandatory according to SI guidelines. I am not suggesting that WP be bound to follow all SI rules, but it certainly should not *forbid* good practice. I disagree with the change, which should be reverted anyway because it was made without consensus. Thunderbird2 15:17, 1 October 2007 (UTC)

Addressing most of this below, but the consensus issue I'll address here: Please read WP:POLICY and WP:CONSENSUS more closely. A change to the wording of a guideline (or anything else at WP for that matter) does not need to establish consensus to make the edit before the edit is made if the edit already reflects consensus. Otherwise, there would be 18,000 discussions per day on talk pages about fixing typographical errors. The overwhelming consensus on how to format decimal fractions on Wikipedia is self-evident simply by reading Wikipedia articles, in which SI notation is virtually totally absent. Per WP:POLICY, which explains that guidelines reflect practice and consensus rather than dictate it, I have simply updated this document to conform to what the consensus already is. — SMcCandlish [talk] [cont] ‹(-¿-)› 23:05, 1 October 2007 (UTC)

If it looks strange to me, and my eyes to a double-take—still—then it's not good for our generalist readers. You think, WTF, squint, stare, and move on (most people won't know what it is). Please give one good reason, apart from your particular familiarity with this practice, that WP should adopt the SI practice in this case. We have a very different readership from that for which SI is aimed, don't we? Tony (talk) 15:23, 1 October 2007 (UTC)

I wasn't arguing for adoption. Rather, that the practice should not be forbidden. Nor should it be discouraged even for articles of particular relevance such as kilogram, metre and second. More generally, SI guidelines often have as motivation the wish (or need even) to minimise ambiguity - a desirable goal for an encyclopaedia. Thunderbird2 15:49, 1 October 2007 (UTC)

SI also says to space groups of three digits before the decimal place, too, doesn't it? WP forbids that, I think. What is the advantage in allowing it in any position, especially since we can't use thin spaces here as we'd like (and which are much more appropriate for this role than normal spaces)—stupid Internet Explorer 6, as usual, screws up the display. It's all very well for SI and hard-copy. Let them. Tony (talk) 16:02, 1 October 2007 (UTC)

Yes, SI uses spaces every 3 digits, both before and after the decimal point. Here there's no need for them before the point because we use commas instead. They still play a valuable role *after* the point though, to make long sequences of digits easier to read, and thus less likely to be misinterpreted. Thunderbird2 16:44, 1 October 2007 (UTC)

I find 456,567,234,123.675,345,789 to be very hard on the eye. Where is the decimal point in all that lot? I find 456 567 234 123.675 345 789 or for Central Europe 456 567 234 123,675 345 789 usage, much easier on the eye. 2007-10-02 11:55 UTC 85.210.15.78

Only valuable if you already know that convention. Otherwise you read it as a string of separate numbers, Less than helpful, it is actually misleading. Rmhermen 22:00, 1 October 2007 (UTC)

You read 234 567 789 123 567 as a list of separate numbers? No. A list of separate numbers would be written as 234, 567, 789, 123, 567 with a comma after each number. So, you want to use commas both as thousands separators and as list item delimiters too? How do you find this 123,345,123, 345,678,234.678,222, 345,789,234, 678,234.345 for readibility then? That is a list of only FOUR numbers. The SI method would use: 123 345 123, 345 678 234.678 222, 345 789 234, 678 234.345 instead. The eye is guided to the commas as list-item separators. The spaces break the numbers into thousands and thousandths. I much prefer that format. 2007-10-02 12:05 UTC 85.210.15.78

Moot: No sane editor here would do any of that gibberish – commas or spaces versions – but would put each figure on its own *-bulleted line, and use typical 123,456.7890 notation, not Continental SI spacing, to prevent any ambiguity at all. Please recall that we are talking about en.wikipedia.org, not fr.wikipedia.org, etc. Central Europe's concerns and preferences are not ours; cf. how we do not even bother to discuss the «angle-bracket-style quotation marks» favored by German and a number of Slavic languages (and probably others; I think the French like them too); they simply aren't relevant at this site. — SMcCandlish [talk] [cont] ‹(-¿-)› 06:22, 3 October 2007 (UTC)

It will confuse our readers. It should be strongly deprecated (a guideline can't actually "forbid" anything). This is Wikipedia, written for a general audience, not SIpedia. As noted above, SI is also the source of nonsense like "5 033 304.72", which we had already long deprecated. We cannot sanely import half of an SI recommendation and then deprecate the other half of it! The spaces' role is not valuable here, because our readers do not recognize it, it will be woefully inconsistent, it is too ambiguous, it cannot be formatted properly (no thin-space support in world's most popular, still, browser), and attempting to simulate a thin space with HTML <font size="-1">&nbsp;</font> markup not only depends upon deprecated code it also makes for wikicode that is harder for average editors to edit. The funky spacing does serve a valuable role in SI documents, because the average reader of them is familiar with and expects the notation, and it is used in circumstances (i.e. print journals and the like) in which it actually works as intended. By way of comparison, consider the standard medical term "myocardial infarction"; in a medical journal, the common-parlance term "heart attack" would be inappropriate and out of place, but in a Wikipedia bio article the exact opposite is true. Wikipedia does not use technical jargonistic terminology and formatting without a very good in-context reason. See also above and in archives; this larger issue has been hashed to death and beyond, in years of debating over MB/MiB, where the closest thing to consensus that appears to have emerged is that we don't force the reader to learn the *iB system unless the distinction between *B and *iB in the article in question is very important. Using SI notation for decimal fractions is not important at all on WP, even in scientific articles, as the meaning of the numbers in question does not change and WP is not a science journal. PS: Thunderbird2's point about an encyclopedia's need to minimize ambiguity is the strongest possible argument against using this SI notation, since it introduces the very strong ambiguity about whether what is being shown is one number or several. I'm with Tony and Rmhermen in doing a double-take every time I see this. And I'm university educated. A 12-year-old isn't going to understand this notation at all. — SMcCandlish [talk] [cont] ‹(-¿-)› 22:48, 1 October 2007 (UTC)

I prefer the SI notation and would have it used everywhere but

• either use it everywhere or nowhere not
  • use it on these articles but not those,
  • use it in these contexts but not those nor
  • use it for these numbers but not those;
• don't use it with an ordinary space (and since IE6 can't hack a thin space, well, that's that) and
• definitely don't use commas before the decimal point and spaces after.

Commas are what we're using (even such functions as {{formatnum: }} are designed around this convention), let's stick with that. But if you think that's bizarre get a load o' pi#Numerical value. Jɪmp 00:24, 2 October 2007 (UTC)

Understood. If even those who prefer SI notation do not like its half-assed implementation in a few places here, I think we can be pretty secure in advising against doing that.  :-) PS: It's not so much that the practice is (objectively) bizarre, it is that it is inconsistent and confusing to the majority of our readers, who are not scientists. Making it internally inconsistent and even more confusing, including to those who prefer it (when done properly), by only adopting half of it, will, as Jimp appears to note, be even worse. — SMcCandlish [talk] [cont] ‹(-¿-)› 04:55, 2 October 2007 (UTC)

In my experience 12 year olds learn quickly and would have no trouble at all with interpreting the spaces for what they are - a device to aid the reader. If there is a problem, it would be with those of us who are more set in our ways. Which of these 3 do you find easiest to read?

• 3.14159265358979323846264338327950288419716939937510
• 3.14159 26535 89793 23846 26433 83279 50288 41971 69399 37510
• 3.141 592 653 589 793 238 462 643 383 279 502 884 197 169 399 375 10

I see no harm in WP embracing this convention for long strings of digits. Perhaps the length of the string would be a better criterion for its use, rather than by article (there I agree with Jimp). Thunderbird2 07:25, 2 October 2007 (UTC)

Well I learned pi to 250 decimal places in my teens (and still know most of it) and that printout was a space after every five digits. So to answer your question - that's what I find easier to read. :-) But seriously, spaces and commas prior to the decimal give you a sense of magnitude of the figure - most important (e.g. 5,058,223,172). Spaces after the decimal aren't that important and a lone figure sitting out on its own after a 7 decimal place figure is just awful (e.g. 3.141 592 6). Most figures in WP are probably under 6 decimal places anyway and most would be less than that. If that's the case - have no spaces. But I do agree when it is long it is better to break them up. But to be honest, by fives or even tens are better than three. Jim77742 07:39, 2 October 2007 (UTC)

I'd prefer threes or even sixes to match the groupings before the decimal point. Which of the three πs do I find easiest to read? The one grouped into threes but who wants to read it anyway? With great long strings of digits like that you'd more likely want to copy and paste it into a spread sheet than actually read it i.e. verbatim. Fives or tens verses threes or sixes ... I call it a moot point on Wikipedia. Jɪmp 08:11, 2 October 2007 (UTC)

Yep, I find the threes easiest to read, but general readers will find it harder to identify for what it is, when filled with these relatively large normal spaces (as opposed to the thin spaces that we can't use). How often do these long strings appear in WP, anyway? Tony (talk) 11:14, 2 October 2007 (UTC)

So rarely that it does not deserve a mention here. At least 99 times out of 100 it is not appropriate to quote more than 3 significant figures, so the question doesn't arise. For the remain < 1%, why not leave it to the editor's discretion? Thunderbird2 11:52, 2 October 2007 (UTC)

I know it's rare. But it happens. And numbers like 3.141 592 6 just look awful when spaced like that (either in text or in an infobox). Spaced out when on a line by themselves with a lot of digits is fine:

3.141 592 653 589 793 238 462 643 383 279 502 884 197 169 399 375 105 820 97

As a resolution can we have no spacing in decimal places except it's optional when the number of decimal places is eight or greater and the number is on it's own line? (A mouthful - can someone word it better?) Jim77742 00:12, 3 October 2007 (UTC)

"Digits after decimal point are written without spaces with an optional exception for numbers with more than seven decimal places on their own line." How's that? Right, but I'd prefer the rule without exception. I don't want to read such numbers, I might want to use them, in which case I'd have to delete any spaces in order to make it one number. Jɪmp 2007-10-03 01:19 01:19, 3 October 2007 (UTC)

Have to concur with Jimp and Tony that no one here is literally going to read a 47-place decimal, so the readability point is moot, and further with Jimp that actually using it, e.g. because you need to copy-paste it into a mathematical calculation, makes the spaced variant user-hateful. Further concur with Jimp (not Jim77742) that one standard is better, and we should not make an exception just for the heck of it; doing so would not even satisfy the mainstream of the SI-supporter crowd, only those who think that using the less significant half of the SI recommendations in this area while explicitly rejecting the other half would somehow make sense to our readers. — SMcCandlish [talk] [cont] ‹(-¿-)› 06:10, 3 October 2007 (UTC)

After reading all this I'd agree with no spaces in decimal places. At all. Although there may be places where it is appropriate (e.g. pi) and better for digit memorisers the cut/paste thing has greater weight for me.

"Digits after the decimal point are written without spaces" Jim77742 07:01, 3 October 2007 (UTC)

I see not one good reason on this talk page that SMcCandlish's addition should be removed again. Tony (talk) 07:04, 3 October 2007 (UTC)

I have no objection to tightening the language of it. It might also be appropriate to invite commentary from Talk:Kilogram since that seems to be the locus of support for SI spacing of decimals. I figure if WP:MOSNUM regulars right now say there is consensus on this, we're likely to have a flamewar result when we fix Kilogram to conform, if the regulars over there do no get to air their (probably already addressed) issues. Procedural point, maybe, but what the heck? — SMcCandlish [talk] [cont] ‹(-¿-)› 07:18, 3 October 2007 (UTC)

I see plenty of good reasons why the addition needs to be more carefully *formulated* before it is added, and I'd like to be given some time to compose my thoughts. In the meantime, Conversion of units is another example where (good) use is made of the 3 digit grouping. Thunderbird2 07:47, 3 October 2007 (UTC)

Just a politics point, but it's usually not helpful to respond with sarcasm to concessions in your argument's direction. That said, there is no delineated time limit here. I think I'm the most "hot" editor here to go change the markup at Kilogram (though not the first to report the issue), so I wouldn't worry too much about the clock, unless you figure to disappear from the discussion for a week all of a sudden. :-) If I were overheated to go change it, it would already be changed; instead, I'm actually the one arguing for more input from that direction, so your metaphoric teeth around my ankles seem a bit inappropriate. I did not mean to imply that no other article anywhere on WP used spacing of right-of-decimal numerals; only that no activism on the part of such spacing has been evidenced by anyone so far that I know of except among the regular editors of Kilogram (frankly, I'd be very surprised to find a significantly different editorship at Conversion of units). So, if you think Talk:Conversion of units people should be invited to comment here as well, then by all means do so. I doubt that any arguments could be presented that have not already been addressed. The gist is, it would make no sense, and would be dreadfully confusing to our readers, to adopt half of an SI recommendation yet reject the other half, and consensus has already very strongly rejected that other half. Arguments in support of adopting the right-of-decimal SI preferences have a very uphill battle. No ire intended at all in that statement; I'm simply stating the situation as I see it. Re: "more carefully *forumulated*": What did you have in mind? Even I've hinted that the new text could be shortened, but I do think that the logic represented, however tumidly, in fact actually and accurately represents general WP consensus. There simply isn't support for SI decimal notation here, however much SI thinks it is wonderful. — SMcCandlish [talk] [cont] ‹(-¿-)› 08:48, 3 October 2007 (UTC)

No sarcasm was intended in my remark. My emphasis on the word "formulation" was intended to indicate that I see the logic in having such a statement in principle, but that I disagree with the choice of words. More explicitly my position is this:

• I agree that short lists of digits do not benefit from the spaces (1.2345 is easier to read than 1.234 5);
• I think that longer lists (like pi, or the speed of light) can benefit because it dramatically improves readability;
• and that the wording of the proposed new text can be improved to say this.

In a nutshell, for a situation in which the SI practice improves readability, I say use it, and if it detracts from readability, then not. Like Jim77742, I remember learning pi to N decimal places (though I can't compete with his impressive N=250!), and like him I did so in groups of 5 digits. The main point is not so much whether the digits are grouped in threes or fives, but that grouping them dramatically improves readability. In doing so, the spaces improve editors' ability to check the correctness of long lists, and therefore improves the overall quality of WP. One point I do not agree with is the copy paste argument. Anyone relying on the accuracy of WP to obtain the value of pi (or any other constant) to 50 decimal places needs their head examined (for that purpose there are vandal-free sites, like this one for pi). The purpose of WP should be to make the list *readable*. Thunderbird2 16:20, 3 October 2007 (UTC)

If we can't rely on WP for calculations, how can we rely on it for memorisations? Indeed, why even read such an unreliable datum? I still value consistancy & useability in calculations over readability. Jɪmp 03:54, 9 October 2007 (UTC)

There are two attributes I value most in an encyclopaedia. These are accuracy and readability. Consistency is important too, but it's not as if this issue arises on every page (if it did I would be arguing for adopting SI guidelines in their entirety). I believe that both accuracy and readability are compromised by removing the spaces. Here's an example I found in Orders of magnitude (numbers), the editors of which write out some very large and very small numbers long hand. The proposed new wording would require them to write the smallest of these (10^-36) as

• 0.000000000000000000000000000000000001

in preference to

• 0.000 000 000 000 000 000 000 000 000 000 000 001

I maintain that the second is much easier to read. Consequently it is also much easier to *check*, thus improving accuracy as well (if you don't believe me, try counting the zeroes). Thunderbird2 08:26, 9 October 2007 (UTC)

Yeah, it's easier to check by counting zeros but you can't do this {{#expr:0.000 000 000 000 000 000 000 000 000 000 000 001}} nor copy and paste it into Excel to check it automatically. As for that particular article, I don't understand why they've written the numbers out long-hand in the first place. Jɪmp 08:47, 9 October 2007 (UTC)

I think the purpose is to explain what is meant by the exponential notation (10^-36). Thunderbird2 08:54, 9 October 2007 (UTC)

Explain it to whom? Do we expect those who don't understand exponential notation to understand SI notation? "A 12-year-old isn't going to understand this notation at all." writes SMcCandlish. Does he not have a point? Jɪmp 23:49, 9 October 2007 (UTC)

I don't see the relavance of age here. It is simply unwise to assume that all who understand decimal notation, will also understand exponential notation. That's why it should be explained. Thunderbird2 08:58, 10 October 2007 (UTC)

The twelve-year-old I intended merely as an example of one of those who might not understand a give form of notation, no, age is irrelavant. "It is simply unwise to assume that all who understand decimal notation, will also understand exponential notation." absolutely but I argue that it's similarly unwise to assume that all who don't understand exponential notation, will understand SI notation. Commas before and no seperators after the decimal point is, for better or worse, the most common form in English (unless I'm mistaken), thus if we're explaining what is meant by the exponential notation (And why else write such numbers out in decimal form?), is this not the way we should go? Jɪmp 08:43, 11 October 2007 (UTC)

Yes, I agree that no spaces after the point is more common. The question is whether this standard use of English should be applied to numbers that appear in only a small proportion of texts (in English or any other language). It is only those numbers with more than 3 decimal places that the issue arises at all. And even I agree that up to 5 decimal places we should follow the standard rule. So perhaps we are talking only about articles quoting decimal numbers to 6 or more decimal places. Regardless of your or my preference, reading the whole discussion over, it seems to me that arguments have been made in favour of the following 3 options:

• insist on no spaces at all (standard English rule);
• insist on spaces every 3 decimal places (standard SI rule);
• insist on no spaces at all for up to 5 decimal places and rely on editors' judgement for the rare situations where more are needed.

Is that a fair summary? I kind of agree with SMcCandlish that all of the arguments have been heard by now. So why don't we just select one of the 3? (by whatever process is considered appropriate) Thunderbird2 09:50, 11 October 2007 (UTC)

Time in astronomy

I suspect that we need to add a bullet point to the Calendar section to cover time in astronomy. For example the articles Julian year (astronomy) and Astronomical year numbering suggest that such articles my need to have exceptions to the general rules in the calendar section. --Philip Baird Shearer 10:32, 4 October 2007 (UTC)

Proposition for the aesthetic format iteration of the date unit of measurement

To assist the Wikpedia online audience with readability, ease of visual cueing, maximizing textual 'texture' for perceptual interest, mnemonic enhancement, deletion of redundant and inefficient separation of space-marker betwixt date proper and unit of measurement (to minimize unnecessary utilization of digital storage resources), and for pure aestheticism: how may I open a voting and/or dialogic process to invite and initate a change in Manual of Style (MOS) guidelines?

• e.g. convention in current MOS: 1008 BCE
• e.g. propositional amendment to MOS: 1008BCE
• this proposition is tendered only in respect of the CE/BCE dating convention and unit of measurement.
  

In my humble and informed opinion, the proposed convention is beautiful when perceived in the context of the uploaded/retrieved page.
Thanking you in anticipation
B9 hummingbird hovering ^{(talk • contribs)} 00:42, 6 October 2007 (UTC)

Sorry, but I disagree. I think the current style is fine. —MJCdetroit 02:04, 6 October 2007 (UTC)

Agree with MJC. Tony (talk) 02:48, 6 October 2007 (UTC)

Note that the proposed amendment requires twelve bytes more than the current convention; the claim that this minimizes storage is patently false. TomTheHand 03:25, 6 October 2007 (UTC)

Double-disagree. Not only are small-caps a typographic convention that is rather archaic and serving no logical purpose (the "aesthetics" argument is entirely subjective), along with numerals like "9" having tails below the baseline like a "g", the usage you propose is simply wrong, from a semantic markup point of view. This would have to be done in CSS with a font-variant:smallcaps, if I remember my more obscure and disused CSS correctly, not with <small> which is not only a misrepresentation of the typographic smallcaps that some publishers use for century and era abbreviations, but is also deprecated in XHTML to begin with. We shouldn't encourage bad coding habits, especially for only subjective reasons of appearance. — SMcCandlish [talk] [cont] ‹(-¿-)› 00:15, 10 October 2007 (UTC)

PS: Please do not wikilink everyday words; it is distracting and makes parsing your text more difficult. FYI: Doing so is also likely to be perceived as an intelligence insult by other editors.; everyone knows what "beauty" and "humble" mean. :-) — SMcCandlish [talk] [cont] ‹(-¿-)› 00:18, 10 October 2007 (UTC)

One would use 1080 <acronym>BC</acronym> or 1080 <abbr>BC</abbr> (and perhaps a class attribute) in HTML with styling provided by CSS, but we cannot expect authors to supply that much markup and the software does not automatically generate it (but perhaps may some day). Small caps is an alternative style of lowercase letters (minuscules), not of uppercase ones (majuscules), so one would indeed simply decrease the fontsize. Small caps are seldomly done right (especially in browsers), because, unlike fontsize alterations, they require extra work by the font designer. Neither small capitals nor text figures are really archaic, though, they are just not as well supported by electronic fonts and DTP software and therefore were hardly used. – Christoph Päper 14:51, 10 October 2007 (UTC)

Found a bug in date formatting (commas between serial dates being swallowed)

To demonstrate the bug, here's an example shown as edited text:

... occurred on [[March 14]] [[1873]], [[April 12]] [[1873]] and [[July 2]] [[1874]].

yielding this displayed text:

... occurred on March 14 1873, April 12 1873 and July 2 1874.

Notice how the desired comma between the first two dates disappears.

This can be fixed thus:

... occurred on [[March 14]] [[1873]]<nowiki>,</nowiki> [[April 12]] [[1873]] and [[July 2]] [[1874]].

yielding

... occurred on March 14 1873, April 12 1873 and July 2 1874.

but that's ugly. (If you want to see something really ugly, look at the edited text for this "fix" in this message!) +ILike2BeAnonymous 07:14, 8 October 2007 (UTC)

I'm totally confused by this. I see no missing comma. And the fact that there's clearly an issue is further reason that the whole date autoformatting thing is a lost cause, at least until it's completely overhauled. Just so many technical disadvantages that it's not funny. Tony (talk) 08:31, 8 October 2007 (UTC)

Just to address your first point (not seeing the lost comma), how are your display preferences set? I use the default (American date formatting). If you're using UK date formatting, change it to the default and look at it; the comma after the first date is definitely being removed. I haven't tried this using the UK version. +ILike2BeAnonymous 18:18, 8 October 2007 (UTC)

Just for the record, at least on my screen there is definitely a comma missing between "1873" and "April". (I checked under Date and Time preferences - I have no preference selected) The sensible solution would be to write it as 14 March instead of March 14. Doesn't the problem go away then? Thunderbird2 18:29, 8 October 2007 (UTC)

No, that would not be sensible at all. I didn't mention that this bug was discovered in an article on an American city. We use American date formatting there; we don't write "14 March". The formatting stuff should work correctly either way, dontcha think? +ILike2BeAnonymous 18:36, 8 October 2007 (UTC)

Well, sure a computer can be taught to recognise and present either format correctly, so I guess the bug can be fixed. But I am used to recognising either "14 March" or "March 14" as valid dates. I would use the former if it were followed by a year (14 March 1873) and the latter if preceded by one (1873 March 14). To be honest I don't see this as British vs American. Isn't it just common sense? Thunderbird2 19:03, 8 October 2007 (UTC)

No, it's incorrect; we (Americans) don't write dates that way. It's that simple. We certainly shouldn't have to change the way dates are written to accomodate some failing in the software here. +ILike2BeAnonymous 19:23, 8 October 2007 (UTC)

Just for the record, to me the original and "fixed" version look identical, both with comma in place. I'm using the ISO preference setting. −Woodstone 20:03, 8 October 2007 (UTC)

Pardon my ignorance, but which one is the "ISO" setting?

Also, at the risk of annoying those who already know this, let me remind all that the vast majority of readers here are going to have no preference setting at all, apart from the default setting, so that setting should work correctly. +ILike2BeAnonymous 20:41, 8 October 2007 (UTC)

I don't understand, why are you removing comma from American dates in the first place (or better yet writing them without commas) :\? Matthew 22:00, 8 October 2007 (UTC)

Writing them without commas won't matter; when they are linked for preferences, the software will add them in Month DD, YYYY format even if preferences are not set. So even if we do have some people like the original poster here who don't follow normal rules, we do get them when we look at the article. Gene Nygaard 15:00, 9 October 2007 (UTC)

Christ, how many times does this need to be said: it does matter. Firstly, Wikipedia articles are forked, if the article lacks commas the forked version will lack commas also (and if they're not using the same setup as Wikipedia (MediaWiki) the commas won't be added on page output). Secondly, commas will not be present when using software like popups. Also the fact remains that we shouldn't tolerate bad grammar in the article's source (basically your argument is: the software fixes it, it doesn't matter… the same argument could work for spellings if the software fixed them, haha.) Matthew 15:32, 9 October 2007 (UTC)

This is weird. I just changed my preferences to the January 15, 1975 version and the result I get in the first listing above is "March 14, April 12, 1873, 1873 and July 2, 1874" with the first 1873 moved over. Gene Nygaard 15:08, 9 October 2007 (UTC)

I believe the date formatting code does in fact expect commas between month-day and year in American format; while it will usually compensate if this is missing, I guess it won't always. Try this version:

... occurred on [[March 14]], [[1873]], [[April 12]], [[1873]] and [[July 2]], [[1874]].

yielding this displayed text:

... occurred on March 14, 1873; April 12, 1873 and July 2, 1874.

— SMcCandlish [talk] [cont] ‹(-¿-)› 00:23, 10 October 2007 (UTC)

Nope there is definitely a bug here; I just tested it and it didn't work. It appears that the <nowiki>,</nowiki> workaround is the only viable one right now. Agree with Tony above that this is more evidence that the date formatting system now is place is royally hosed. — SMcCandlish [talk] [cont] ‹(-¿-)› 00:26, 10 October 2007 (UTC)

A semicolon to separate elements which themselves contain commas would fix it and would actually be more correct and easier to read for the March 14, 1873; April 12, 1873 and July 2, 1874 format (the OP didn't include the serial comma after the second 1873), but it would leave you with a semicolon for elements not containing commas in 14 March 1873; 12 April 1873 and 2 July 1874. Gene Nygaard 06:38, 10 October 2007 (UTC)

Level of precision

I have made the following changes.

old

Converted values should use a level of precision similar to that of the source value; ... The exception is small numbers, which may need to be converted to a greater level of precision where rounding would be a significant distortion; ...

new

... The exceptions are values with only one significant figure, which, to avoid the introduction of inaccuracy, may need to be converted to a greater level of precision; ...

• Technically speaking they are values rather than numbers.
• It doesn't matter about the size of the "number" (What does this even mean? Which is smaller 0.01 millilitres or 10 nanoseconds?); what's important is the significance of that value.
• It's perhaps a little rough and arbitary but one significant figure seems a fair place to draw a line for general purposes. If you've got at least two significant figures, you've got at most ±5% inaccuracy.
• Rounding to a greater level of precision is still rounding.
• The rounding is not the distortion but that which introduces the distortion—an admittedly somewhat picky grammar point.
• I feel that inaccuracy is a better description of what we're dealing with here than distortion.

Jɪmp 07:28, 9 October 2007 (UTC)

It is probably important to understand that the precision of a number isn't necessarily determined by looking at any one single measurement, when similar related measurements are included in the same article—or when we are otherwise aware of the precision normally used in the context. The trailing zeros before the decimal point are sometimes significant digits (and normally are when they come after the decimal point, except sometimes in dollars and cents figures and the like). Gene Nygaard 14:55, 9 October 2007 (UTC)

This is true. Should we work it into the text or can we assume editors should be aware of such considerations? Of course, this was as much a problem with the old text as with the new since it's a detail that neither really address. Jɪmp 23:40, 9 October 2007 (UTC)

I'd leave it and avoid instruction creep. Some will know it. The proliferation of conversion templates, which most editors probably won't understand when it comes to the nuances of setting precision (and which employ a number of different methods of doing so, so even those who know it is possible might get it wrong by trying the wrong method) can be a problem in this regard, even if they do provide more formatting consistency and more accuracy (if the right template is being used in the case of ambiguous units). Gene Nygaard 06:45, 10 October 2007 (UTC)