Wikipedia talk:Manual of Style (dates and numbers)/Archive 29

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 25 Archive 27 Archive 28 Archive 29 Archive 30 Archive 31 Archive 35


Following on from an archived discussion on currency conventions, I'm hoping to nail down some specifics on dealing with currency: comments are welcome at Wikipedia:Naming conventions (currency). Ziggurat 02:17, 28 September 2005 (UTC)

Have been bold and merged it here per merge tag. Hiding talk 20:02, 9 November 2005 (UTC)

This is a placeholder while I can come up with an eloquent way of stating that there needs to be a Wikipedia-wide convention on how to indicate currencies (US dollars vs. Australian dollars, etc.) and of which symbols to use (no one seems to know what to do about the Yuan). --Dante Alighieri 01:17 17 Jul 2003 (UTC)

I suggest we use the ISO standard 3-letter currency codes defined in ISO 4217. This would have a number of advantages:

  1. the system already exists and is standardised
  2. we don't have to invent/decide which name to use, eg. the Renminbi is CNY
  3. every currency has a unique code, which prevents ambiguity

-- Cabalamat 16:28, 3 Sep 2003 (UTC)

The trouble is, most people don't know the codes. "2000 chinese renminbi" is more understandable than "CNY 2000" DJ Clayworth 22:00, 21 Nov 2003 (UTC)

In general, we defer to whatever is common practice for published English works in the English encyclopedia, why not do that here? It's probably reasonable to expect almost all English speakers to recognize the U.S. dollar, English pound, and Euro symbols, so just use them. Otherwise, the unit is usually just written out ("one million Japanese yen"). What do the style guides suggest?? I can imagine "$1 million (Australian)" being fine. If there's a need for an abbreviation (because the unit repeats throughout the article), use the abbreviation from ISO 4217, but define the abbreviation inside the article before using it. If you use almost all the abbreviations, at least cross-reference to the ISO page. -- Dwheeler

While I am normally a big fan of following ISO, this is one case where use of the standard is simply contrary to good English style. A good example is the Egyptian pound; ISO 4217 calls for EGP, which is virtually never used in Egypt (except maybe at foreign exchange counters). There are two commonly used abbreviations: in Arabic it is simply the letter gim (for guinea), and in English and other European languages it is L.E., for livre(s) egyptienne(s). In an article dealing with Egypt in English, the only acceptable abbreviation would be L.E., which in Wikipedia should be defined on first occurrence in an article. Similary, GBP should be £, JPY ¥, EUR €, etc. —Tkinias 16:55, 26 Mar 2004 (UTC)

I am absolutely against using abbreviations except for the most commonly known symbols, as Tkinias suggested. I think there should be a preference for spelling out the currency -- 1,000 U.S. dollars (although $1,000 is also perfectly acceptable), 1,000 pounds sterling, 1,000 Australian dollars, 1,000 euros, 1,000 Japanese yen, etc. acsenray

If $1,000 is also perfectly acceptable then €1,000 is also acceptable. --Gangulf 08:11, 27 Sep 2004 (UTC)
Shouldn't the symbols for sterling (£) and yen be acceptable... as they are practically unique to their own respective areas... the dollar symbol is much more widespread.

"$1" would confuse me. Wikipedia lists over 30 currencies that use "$" (here). Only USD uses the S-with-two-bars symbol, but that's hard to type.

The ideal solution is the same as the Wikipedia date standard; encode the ISO amount in the text, and have Wikipedia software display the article in a commonly understood format containing a link to the currency's definition. (E.g. "[[USD]]" would become equivalent to: "[[USD|$]]" - which Wikipedia would render as S-with-two-bars.) Although this envolves developer work it has the advantages of:

  1. Readability & Comprehension where the ISO symbols are obscure.
  2. Lack of ambiguity, and other advantages of ISO (listed at the top of the page by Cabalamat).
  3. Ability to customize, and therefore standardize, the way currency amounts are shown. (Like the date preferences.)
  4. Makes it easier to insert exotic currency signs, which make Wikipdeia prettier.
  5. Deals with the technical problem of rendering unicode symbols in all browsers (probably through user preferences to display ISO-4217 where unicode is mangled).

If ISO 4217 is not used, the convention must specify an efficient way of discussing the AUD/USD exchange rate, old Turkish Lira vs. new Lira, and prevent confusion of currencies with similar names. Wragge 16:19, 2005 Apr 17 (UTC)

On the dollar issue: surely a good idea (especially for the dollar) would be to use, for example, US$, AUS$ and Z$ or ZIM$ (although ZIM$ is more explicit, I have never seen it. Z$ seems quite common). I use R 0,00 for rands (my home currency), but recognise ZAR (ISO code) in international contexts. Same with british pounds (GBP). However, maybe a new notation like [[ISO-code]] is better. It makes hard-to-type (like pounds, euros and new sheqels on my keyboard) much easier.

User:Taejo made the above comment, by the way. I now think that there is an even more important benefit of the new notation: Converting amounts. A page I was editing yesterday lists amounts in lac, in the format "Rs 29 lac"; not being used to the Pakistani exchange rate I really don't know how significant that is, and to be of much use to the average reader the amount must be converted into a well known currency - presumably in the (configurable) format: Rs 29 lac (24 million USD).
Naturally this raises many issues about which exchange rate to use (PPP vs nominal, and which year's rate to use) but any choice would make an unintelligable amount into something which could be immediately understood (as an order of magnitude).
Any thoughts? Wragge 08:55, 2005 Jun 1 (UTC)
Hmm, interesting... yes, having a standard certainly would make it easy to do auto-conversions. Maybe someone should make a dev request? ;) --Dante Alighieri | Talk 20:53, Jun 2, 2005 (UTC)


Style not naming convention?

The same issue came up in this discussion: Currency Style Dismbiguation. Surely this is a Style question rather than a naming convention? (i.e. it refers to the formatting within articles rather than how they are named). There should definitely be some kind of style guideline for currency. NPOV makes frequent reference to the international nature of Wikipedia (see especially Wikipedia:Neutral_point_of_view#Anglo-American_focus), and yet despite twenty three countries using the dollar there are several articles in which it is implicitly used solely for the U. S. dollar. In many articles of a specifically national nature this is not an issue (that it is US$ is obvious from the U. S. context of the whole article), but there is a clear NPOV bias in comments like this from 2004: "The People's Republic of China to invest $20 billion dollars in Argentina, a deal signed days before the Asia-Pacific Economic Cooperation forum to be held in the City of Santiago in Chile." Where more specific terms are used there are several styles throughout Wikipedia, so it needs to be standardised.

Of the major style guides, the APA and the MLA say nothing, and the Chicago Manual of Style discusses it briefly but only mentions one example: "In Canada the current quotation was $2.69 (U.S. $2.47) a pound." The MHRA Style Guide recommends using an appropriate abbreviation before the symbol (C$/Can$, A$/Aus$, NZ$).

International newspapers tend to display the same Americentric bias: Guardian "abbreviate dollars like this: $50 (US dollars); A$50 (Australian dollars); HK$50 (Hong Kong dollars)" Times "with figures use $5 (when American), A$5 (Australian), C$5 (Canadian), S$5 (Singapore) and so on"

The World Bank style guide World Bank (pdf) uses "U.S. dollars or US$"

In Wikipedia, most articles already use some variation of the style suggested in the international papers, MHRA, and World Bank, for example:

But some pages with an international focus need to be standardized:

"For example, GDP per capita in China is ca. 1,400 U.S. Dollars, while on a PPP basis, it is ca. 6,200 US$. At the other extreme, Japan's nominal GDP per capita is ca. 37,600 US$, but its PPP figure is only 31,400 US$"

"The Educational Testing Service (or ETS) is the world's largest private educational testing and measurement organization, operating on a annual budget of approximately $900 million. ETS develops various standardized examinations primarily in the United States, but they also administer tests such as TOEFL in most nations. "

"The Bank of Credit and Commerce International (BCCI) was a major international bank founded in Pakistan in 1972. At its peak, it operated in 78 countries, had over 400 branches, and claimed assets of $25 billion."

I would suggest a few things:

  1. use the convention "US$100", "A$2000", "U. S. dollars", "Australian dollars", never "$200 (US)" "$US200".
  2. Keep the ISO 4217 link in the articles on the currency only, and if necessary (it's overkill to link every single currency symbol used) link to the appropriate currency.
  3. write it into the Style Guide Wikipedia:Manual of Style (dates and numbers)
  4. establish a policy on when to use $ (i.e. in strictly national articles in which there is no possibility of monetary ambiguity?).

Ziggurat 02:11, 28 September 2005 (UTC)

Units of mass in precious stones

(Discussion moved from Talk:Diamond)

I think the article will be improved by having specialist weight units (carats) supplemented with non-specialist weight units (grams). I made edits to that effect but the article was reverted to remove gram values. See:

This topic was discussed previously at: Perhaps we have drifted apart from what was said in that discussion. I would like to find a solution. Can we go through it again please? Bobblewik 23:07, 28 September 2005 (UTC)

Aside from the simple fact that diamonds are not sold by the gram, converting every carat value to grams does (IMO) impede the flow of the article. The very first mention of "carat" in this article is both wikified and its value converted into kilograms; this gives a clear indication that it is a unit of mass, and if further explanation is necessary, the reader can click through to the carat article. Furthermore, the Diamond#Carat section gives an in-depth explanation.
As for the "specialist" nature of the carat: it's now common for would-be gem purchasers to do a little research before buying, and the term carat has long since entered the lexicon of the general public. It's used in advertisements, TV shows, mainstream movies, and literature. Most jewellery stores (e.g., Peoples, Birks) have informative posters and signs next to display cases to educate the unfamiliar, of which there are increasingly few. As you yourself said in the May 2005 discussion linked to above: "When I first discovered that the carat was simply 200 mg, I was astonished that it was so simple a conversion." (Emphasis mine.) A simple conversion needn't be invoked each and every time. I don't know where in the world you are, but consider that a great many Americans (who make up a sizeable fraction of our readership) are unfamiliar with the gram itself, but that doesn't mean we should convert every g/kg value to ounces and pounds. I think the article is fine as it is. -- Hadal 03:48, 29 September 2005 (UTC)
I pretty much totally agree with Hadal; the conversion is informative the first time it is made, and useful in situations where bulk mining data are mentioned, but beyond that it is pedantic and interferes with the flow, without adding anything. This is not like giving the height of mountains, where the conversion is useful every time; it is more like giving the translation of a foreign name. Our article on Japan, for example, notes that natives actually call it Nippon, but doesn't then go use the construction "Japan (Nippon)..." every time the word Japan is used. - Bantman 18:50, 29 September 2005 (UTC)


We've already standardised date format. It seems to me strange that any time format is allowed given that it is 24-hour, and, furthermore, that the manual actually specifically says not to edit an article just to change the time format (which, as you could probably tell, was exactly what I was about to do).

Like dates, there are lots of different time formats. And like dates, it would be better if they were all the same. I don't know about regional formats though.

Anyway, there are two things in time formats that annoy me when I see them, and I'd be very surprised if others here disagreed with me:

  1. When people write 3.25. Times don't use dots, I thought — it was always meant to be a colon, like 3:25.
  2. Inconsistency in using AM and am and a.m. and A.M.. I know the manual says to avoid 12-hour times (a principle which I oppose), but anyway, this inconsistency is blatant (unlike numbers as words). So this should be standardised. Personally, I prefer lowercase. Don't mind about dots, as long as it's consistant.

There is of course one other thing, that is the thing about 12-hour and 24-hour clock which I understand is very controversial and probably won't gain consensus, though if anyone is like me and wants it to be 12-hour in prose (not technical or tables), please drop a word.

Any thoughts? (If I don't post something for ages it means I've been caught up elsewhere, I'll try and check here as often as I can) Neonumbers 08:47, 3 October 2005 (UTC)

I have a brilliant idea. Let's create a robust template to handle times. Something that would take a time and zone parameter. For example, an instantiation could look like: {{time|2:17 PM|EST}}, which would output 2:17 PM EST (19:17 UTC). I don't know how hard it would be to do that, but the PHP code to convert the times and uniformly case the PM/AM would be incredibly simple. We could then use a bot to snarf up all the times and formatthem uniformly. Local time zones in which 24-hour clocks are preferred would display 24-hour time, and time zones in which 12-hour clocks are preferred would do that, but the UTC time would always print. If a time zone is unknown, the editor can pass LOCAL as the parameter, and the 24- or 12-hour format will be preserved. --Mm35173 19:06, 11 October 2005 (UTC)
Nice idea, but I'm sceptical about its feasibility — the date convention (I know it's being debated) has worked well for a while now; I very rarely see a date that hasn't been correctly formatted. So I don't think a template's the solution in this case because it's a number format and I highly doubt a template would become common use.
My concern is that the current guideline is a non-guideline and, like dates, times should follow a consistent format. This is one of those things you can flip a coin for — it doesn't matter that much so long as they're all the same (but I don't actually want to pick one completely at random).
If I were to suggest that the format (say) 3:45 p.m., 12 noon, 12 midnight, 12:34:56 a.m. for 12-hour and 23:45, 13:42:36 for 24-hour be standardised, wouldn't that be useful for a manual of style? Neonumbers 06:48, 14 October 2005 (UTC)
Tell me, Mm35173, if somebody is trying to show that when an event took place, the streets in that residential neighborhood were quiet because it was 4 o'clock in the morning, local time, exactly what good would it do me to know only what time that was in my time zone? Or what time that was in UTC? Or both? Maybe the idea isn't so brilliant after all. Gene Nygaard 16:13, 14 October 2005 (UTC)
To partially answer my own question, I suppose we could change your "unknown" to "either unknown or the local time is the relevant time", but who are we going to get to even read the instructions on how do it by the time we add all the possibilities, let alone follow those instructions. Gene Nygaard 16:20, 14 October 2005 (UTC)
Maybe if I keep at it long enough, I'll get to my real point. The notion that there are "Local time zones in which 24-hour clocks are preferred" is what really has me buffaloed, and it is whatever you might come up with for a replacement for that wording that is likely to be the greatest complicating factor by far. Gene Nygaard 16:24, 14 October 2005 (UTC)

There are exceptions to any rule. The Hebrew calendar calculates its dates based on times in a 24 hour clock that begins at 6 p.m., because the Jewish day begins at sunset and 6 p.m. is the average time of sunset over a whole year. This must not be converted to 18:00, because 18:00 in the Hebrew calendar means noon. Actually, even noon should not be written as 18:00 because in the Hebrew calendar an hour is divided into 1080 parts (halakim), not into 60 minutes. Noon should be written 18h—a typical Hebrew calendar time might be written as 5h 204p. — Joe Kress 05:21, 19 October 2005 (UTC)
Okay... well, I'm just going to steer this discussion back to its intention (or at least my intention)... For times that use the (Western) 12-hour or 24-hour, 60 minutes to an hour, 60 seconds to a day, format, with am and pm for 12-hour but not twenty-four hour; and this includes localities such as France that (feel free to correct me) use the 12/24-hour clock but not am and pm, preferring instead to use morning, afternoon and night... well anyway, times that follow the 12/24-hour clock:
We pick a consistent format. This is things such as, use colons not dots (or dots not colons), use a.m. not am nor AM nor A.M. (or whatever version we choose), do/don't use a leading zero for the hour (or, use a leading zero for 24-hour but not 12-hour), things like that.
And, this consistent format in place, we allow the editing of articles (by "allow" I mean, it should generally be okay to, I do not imply that this manual is law) to change the time format to conform to this.
I mean, let's face it, other time systems are almost never used in an English encyclopedia and we can't possibly cover all of them here. The people involved in those areas can work something out — but the 12/24-hour clock format is a big issue that concerns just about everyone. So, just the 12/24-hour clock.
Any thoughts on this? Neonumbers 10:41, 19 October 2005 (UTC)

There having been no comments for the past while, I'm going to propose something.

Context will determine whether the 12-hour or 24-hour clock should be used.

Times in the 12-hour clock should with a colon(s), and lower case "a.m." and "p.m.". These suffixes generally cannot be omitted. "12 noon" and "12 midnight" should be used instead of "12 a.m." and "12 p.m."; some readers find the latter ambiguous.

24-hour clock times follow the same format, except without the a.m./p.m. suffixes. Discretion may be used to determine if leading zeroes should be used. 00:00 refers to midnight; 12:00 refers to noon.


12-hour clock Not 24-hour clock Not
2 p.m. 2pm 14:00 14.00
2:34 p.m. 2.34 PM 14:34 1434
12:04:38 a.m. 12.04 38″ A.M. 00:04:38 or 0:04:38
12 noon 12:00 p.m. 12:00

I foresee a debate that takes the lines of "no consensus"; "no discussion", and so on. I waited one week from the last message and there were no replies, and three weeks from the first I posted to which there were no replies answering my question — so the best I can do is propose this change and see who objects. If no-one has strong objection reason then I think it's reasonable to assume people are okay with it.

It is not worth debating, in my opinion, which form of AM am A.M. a.m. is better because let's face it, none of them are "better" than any other and a simple plurality vote could decide it just fine; I'm sure many won't even have a preference.

If there are no objections for ten days, I will replace the current "Time formatting" section with the text above on 5 November 2005. Neonumbers 06:59, 26 October 2005 (UTC)

Just a note that I used the template p.m. = p.m. above which is the same as another user's <small>PM</small> = PM. — Joe Kress 02:54, 28 October 2005 (UTC)
Note acknowledged, but I didn't think it would be worth using a template or such formatting every time a time is encountered, even less so when it's not widespread and general practice. (Don't worry about current occurences of it, they'll fade with time.) Neonumbers 22:48, 28 October 2005 (UTC)
Done. Neonumbers 11:09, 7 November 2005 (UTC)
I'll wade into this too!
Since a.m. and p.m. are originally from the latin, ante meridiem and post meridiem and since there is already a style for the use of latin words, namely that they are italicised et al, etc . Therefore 12:01 AM would become 12:01 a.m.. I don't really care about the italics but the . are important.
Also, regarding the format Suggested by Neonumbers, I believe there should be a differentiation between the minutes and seconds delimiter. As you have written it above 12:04:38 where it should possibly be 12:04.38 to differentiate between minutes and seconds. Definitely keep the colon between hours and minutes, but a period between mins and seconds. DanF 22:44, 7 Nov, 2005
The three style manuals that I have, Chicago, Oxford, and MLA, all call for common Latin abbreviations to be set in roman, including a.m. and p.m. —Wayward Talk 12:05, 7 November 2005 (UTC)

Okay, all :-). I'm reluctant to change the min/sec delimiter to a dot just yet, as I always thought convention was with a colon for both; but if anyone else has advice on this matter, that would be good too. Neonumbers 11:27, 10 November 2005 (UTC)

ISO 8601 specifies a colon as a delimiter between minutes and seconds. We need to reserve use of the period as a decimal indicator. If I see a 1500 metres race result of '4:21.5', it means 4 minutes, 21 and half seconds. Similarly I interpret '10.25 min' as 10 minutes 15 seconds.
I prefer 24 hour format and think it is widely accepted in most of the world. It is particularly useful for things such as indicating opening times and 24 hour operation (e.g. transport) times. In any case I have been replacing upper case 'AM' with lower case 'am'. I don't mind if it is 'am' or 'a.m'. It would be nice to be have a convention on this. Bobblewik 12:15, 10 November 2005 (UTC)
Okay. Use 24-hour format wherever it's appropriate, for 24-hour operation like transport times I'd say it's pretty appropriate. The convention henceforth is "a.m." for the reasons mentioned above by others (and my arbitrary decision as the original proposer.)  :-) Neonumbers 05:39, 13 November 2005 (UTC)

Proposed Currency Section

Following on from the page on currency style - Wikipedia: Naming conventions (currency) - I'd like comments on this proposed section, to be added to the end of the page:

For currency, use the appropriate symbol (before the quantity) or name of the currency unit (after the quantity), for example:

  • "$100" or "100 dollars" not "100$"
  • "€100" or "100 euros"
  • "¥100" or "100 yen"

For national articles or those where the type of currency is unambiguous it is not necessary to denote which currency unit is being used. However, it is necessary to indicate the nationality of currencies in internationally oriented articles or articles referring to more than one currency of the same name. This is done to avoid confusion and potential ethnocentrism. For example, "US$100" or "100 U. S. dollars" (not "100 US$", "$100 (US)", or "USD 100"); "A$100" or "100 Australian dollars."

Standard abbreviations are given at the article on the currency, for example:

We need a guideline, but I'm not happy about the automatic dismissal of the ISO 4217 codes which are the only unambiguous method we've got for indicating currencies. B$ = Bahamas, Barbados, Belize, Bolivia or Brunei? Are formulations such as "USD $100", "$333 (CAD)", or "MXN $250" really unacceptable? Let's explore all the available alternatives. Hajor 00:58, 4 October 2005 (UTC)
I'm happy to explore the alternatives, that's what the talk page is for! :-) I'm suggesting this method as
  • the most common standard in current Wikipedia articles
  • the most common standard in international newspapers (Guardian, Times)
  • the World Bank standard
  • the 'cleanest' technique
That said, I have no objections to the alternative styles you list; if we're going for a standard-to-all-articles style I still favour my proposed technique, but if we're going for an article-consistent style instead they're fine.
Ziggurat 01:20, 4 October 2005 (UTC)
In full support for the existence of this, for standard-to-all-articles. That said...
  • I thought €, by convention, was written after the number, e.g. 45€, and for decimals e.g. 44€99. Or is this just regional differences? I don't know for sure, but that's what I thought so I thought I'd bring it up, feel free to correct me.
  • I prefer US$ to USD, because it's in more widespread use. "USD", like many other international "standards" (like SI units, for example) should only be used in specialised or technical contexts.
  • There is a major downside to using lots of different forms, that is that it looks messy; even if they are consistent within the article, their inconsistency within the encyclopedia degrades the encyclopedia's appearance.
  • I don't want to halt the exploration of alternatives before it's happened, but things like $333 (CAD) and all the other ones are, well, shall we say I've never before seen that format in my life.
And, I have a suggestion:
  • Generally, use the US$ format. If there is risk of confusion with one currency only (like B$), specify at the top of the article in italics ("In this article, B$ refers to Bahamas dollars"). In certain contexts, it may be better to use the USD format; if this is chosen it must be applied to the entire page and the first appearance of each must be linked to the appropriate page.
By certain contexts, well, I don't really know what I mean, I guess I mean in tables where there are lots of different currencies at it's completely unrealistic to specify each one so therefore we turn to our unambiguous standard. Also, in specialised articles that require it, but I don't know what articles would as I have no expertise whatsoever of the area. But anyway, in normal prose USD should be avoided because US$ is more common and understood.
Just to clarify myself, I don't see this as a compromise between standard and non-standard, I am completely against pitiful compromises on manuals of style because it is compromises that make them fail. This suggestion is what I think is best for good style with what I know at present (which may be very little). Neonumbers 10:16, 4 October 2005 (UTC)

Interesting replies. Has anyone got a link for those codes as used by the World Bank? Googling for "currency codes" on site only gives me results that use ISO 4217. I actually suspect there isn't a canonical list of those codes that covers all the world's currencies, and that once you get past the better known dollars, most of them seem to be pretty ad hoc.

Which leads into my next comment. What's been said so far has mostly focused on dollars, pounds, yen, and euros. And I rather suspect that's the case with the two style guides (Times and Guardian) -- they're very centered on the northern industrial world. If the aim here is to set a global standard (gulp), then we need to consider the full range of world currencies -- dollars and pounds, sure; but francs and lempiras and dinars and rupees, too. Cast the net a whole lot wider.

On reflection, perhaps this is something that needs to be taken on a case-by-case basis. Maybe I wasn't casting my net wide enough, either. Those examples I gave above really only work well with currencies that use the £ or $ signs (does anyone else use ¥ besides the Japanese?). For instance, saying

doesn't sound too bad, but it would be frankly bizarre to say

Much better in those cases to use "...costs L.E.0.75" or "costs GTQ 3.50."

(One downside of the ISO Codes is that the standard recommends they be used without $, £, etc, but that's very difficult to do in everyday texts such as these encyclopedia articles -- it feels very unnatural, counterintuitive to write a dollar amount without including the dollar sign.)

Another alternative -- we are a wiki, after all -- would be to use neither system of code and simply to rely on a link to the currency from the symbol to make it clear what we're talking about: $100, £100, 100. That would make the text flow more neatly, but it would have one major drawback: including the ISO 4217 three-letter code makes it crystal clear whether you're talking about ARS or ARP pesos, RUR or RUB rubles, MXP or MXN pesos, ALL or ALK leks, etc. That's lost if we just use the naked symbol, without the ISO clarification.

One thing I don't like about those US$, Z$, RD$, codes is they don't always distinguish between the marker that's used in the country (such as Brazil's R$) and the combination of the domestic marker and additional identifier (such as US$, C$, etc.), in the sense that that's not what's printed on the banknotes or what people write on their checks -- ie, for international consumption only. Thus: the Eastern Caribbean Central Bank's officially sanctioned symbol for the East Caribbean dollar is "$". EC$ is just some form of disambiguation they use internationally to avoid confusion. Also, from an aesthetic point of view, some of them -- even govt. sanctioned ones, such as Barbados's "Bds$" -- are frankly hieroglyphic.

Sorry -- came out a bit rambling, the above. Essential summary: US$, A$, etc. are good (and internationally recognized) for disambiguating one dollar from another, but the system breaks down once you're past that first group of familiar currencies: Is there one for the Uruguayan peso? Is a similar set of symbols used with the pound sign? (I couldn't think of any xx£ examples.) ISO codes are utterly unambiguous but -- as noted above -- not that familiar to non-specialists not accustomed to the world's more arcane currencies and (as also noted) not that easy to work into a grammatically flowing sentence. It may very well be that one system is appropriate for the world's most familiar currency, but not right for the kwanzas and the like that no one's ever heard of. Let's discuss this some more. Hajor 15:14, 4 October 2005 (UTC)

Some of the proposals above look really bizarre to me: USD $100, $333 (CAD), MXN $250, $250 (CLP), EGP L.E.0.75, GTQ Q3.50. They double up the name or even give a wrong name (like mixing $ and peso). Written as USD 100, MXN 250, CLP 250, EGP 0.75, GTQ 3.50 they would make perfect sense. This is the form described by the ISO 4217 standard. −Woodstone 20:32, 4 October 2005 (UTC)
The World Bank standard (pdf) says US$ or U. S. dollars, but for all other currencies suggests that you should "consult Lotus notes" for the accepted abbreviation, whatever that is. Regarding this: "the Eastern Caribbean Central Bank's officially sanctioned symbol for the East Caribbean dollar is "$". EC$ is just some form of disambiguation they use internationally to avoid confusion. " New Zealand is the same; $ is the symbol here (of course) and NZ$ is appended when talking internationally. Ziggurat 20:41, 4 October 2005 (UTC)
You're right about the standard prescribing "USD 100" etc (without the symbol) but, as I noted above (and I quite understand if you missed it), it's very difficult to get lay writers, in lay contexts, to overcome the urge to indicate a dollar amount with a dollar sign. Problem with (eg) "AUD 100" is, outside a specialised context, it's not intuitively a currency amount. Hence the "doubling up". Of course, that may very well be the best argument against using the ISO codes at all. Or one in favour of "CAD 100" on first use, $100" on second, and "$100" subsequently.
Re "mixing $ and peso" -- $ is the symbol used by all the world's pesos, plus a couple of others (the boliviano and the real, for instance). Yeah, our dollar-, pound-, and euro-wielding editors might not be aware of that, but protecting them from having their horizons involuntarily expanded doesn't seem to be a good reason to deprecate standard local use. Hajor 21:02, 4 October 2005 (UTC)
Furthermore, it was a peso sign before it was a dollar sign. Gene Nygaard 16:15, 6 October 2005 (UTC)
I wonder if a full reliance on links (e.g. $100) would be such a good idea — it's still an encyclopedia, and you can't really see links once you print the article (with a printer).
That aside, I'd like to just resuggest my earlier suggestion, because I didn't read anything about it: where necessary, put "In this article, NZ$ refers to the New Zealand dollar" at the top of the article except obviously not with New Zealand because that's unambiguous and therefore unneccessary. Only where necessary.
I think that that would be a good solution because it removes ambiguity without sacrificing understandability. Like I said before, it won't work where there's lots of currencies on the same page, in that case use those ISO codes no-one knows about. If there's really so many currencies it's better to revert to ISO codes, then I figure it'll probably be techincal/specialised anyway.
Of course, I don't claim by any means to be a currency expert, so if that's the stupidest solution ever please say so and explain why. But my point: ISO codes are not ideal and while their are plenty of situations where they'd be of good use, in articles where only one or two currencies are referred to I make my suggestion (as above). Neonumbers 07:00, 6 October 2005 (UTC)
Actually, as far as I know, ISO 4217 does not prescribe USD 100, MXN 250 etc, but 100 USD, 250 MXN etc. I would myself never hesitate about writing "100 USD" or "800 SEK" or the likes in an article. Maybe it's because I'm European and we are more used to international standards than Americans are, I don't know, but I really can't see how anyone, anyone, can read "100 USD" and not immediately think "a hundred US dollars". I feel really strongly that if a common currency symbol (most often $) is used throughout an article to mean the same currency, then say so at the top if it's not obvious, and that's fine, but never ever use the $ symbol if the context needs disambiguation. I don't like US$, and I certainly don't like "$100 USD" which clearly means "a hundred dollars US dollars". And we need the ISO codes for non-dollar currencies anyway. There are for instance at least seven countries sharing the currency name "krona", so the "100 yen" trick won't work here, and there's no symbol to denote it.
As I see it, the only reasonable alternative to the ISO codes, when there are several currencies involved in the same article, is writing out the entire currency each time. "There was a transfer of 800,000 Swedish kronor (roughly 100,000 US dollars) to the establishment, which added an additional 50,000 Euros." Something like that. But I think that's unnecessarily obtuse. Use the ISO code, and link it to the currency. Anyone will understand USD and EUR right away, and if someone is dumbfounded by SEK, he can just hover over the link. (This has the additional benefit of not having to use "kronor", a Swedish plural, in English, or contemplate translating it to "crowns".) -- Jao 07:53, 6 October 2005 (UTC)
I prefer US$125 in most situations, linked on first occurrence.
I find 125 USD acceptable in some situations.
For an isolated occurance, 125 United States dollars is okay.
I hate £3.75bn for £3,750,000,000 or A$1.45m for A$1,450,000.
Even worse is Rs 2.5 crores for Rs 25,000,000, or Rs 47 lakhs for Rs 4,700,000. Gene Nygaard 16:52, 6 October 2005 (UTC)
We appear to have run out of steam without reaching anything close to consensus. Does anyone want to try a watered-down proposal for language to include: perhaps Gene's first line for "familiar" currencies and his second line for "unfamiliar" ones (except there'd be a large grey area where "familiar" is inherently tinged by geographical bias and hence POV). Or should we continue to take matters in each article on a case-by-case basis? One guideline that should be included, and one on which I imagine consensus would be a given, is the use of lower-case letters for currencies: euros and francs, not Dollars and Rupees. –Hajor 16:44, 11 October 2005 (UTC)

Skatebiker 07:35, 12 October 2005 (UTC) I think that using currency can be rather relaxed by e.g. using lower case names, ISO 4217 three letter codes, but e.g. US$ 100 os OK as well. But Z$ can be confusing, use rather ZW$ for Zimbabwe dollar.