Wikipedia talk:Manual of Style/Dates and numbers/Archive 23

'''This discussion is about the guideline on units and its wording. It continues on archive 24.'''

Permitting removal of metric units: inconsistency in the Manual of Style
This is what the Manual of Style says about metric units:
 * Metric units are always acceptable
 * if non-metric units are always given there is usually no need for metric units
 * give the metric equivalent as a courtesy
 * should not be used if they reduce the flow of a sentence
 * should not be used if they interfere with the quality of the writing
 * should be used where these are appropriate
 * multiple equivalents including metric are cumbersome and shall be avoided.

Those phrases are inconsistent, ambiguous and subjective. An editor wanting to remove metric units can claim that removal is consistent with the Manual of Style.

Some of the phrases were added without discussion. Does anybody have any proposals for different wording? Bobblewik (talk) 12:43, 11 July 2005 (UTC)


 * I'm thinking about this as I write, so I'm writing this on the spot. The ideas that we're trying to convey are (respectively):
 * Metric units should not be changed to non-metric at the expense of the metric
 * Some contexts will not require metric units
 * In others, metric units should be there, i.e. length
 * In sentences like, "he was miles from the station", there is no point
 * Same idea as the one above
 * I don't understand that part
 * Don't get too carried away with equivalents
 * So I think the contexts that don't require units are meant to be scientific contexts where non-metric units are normally used, like say astronomy, the AU, and imperial units that the reader will visualise as he reads should be converted to metric. Either that, or the statements are contradictory.
 * I was going to start writing an initial proposal for different wording here but I'm too tired to - I hope these ideas clarify things for a starter. Neonumbers 13:40, 12 July 2005 (UTC)
 * I don't see all of what Bobblewik does, but I do think the topic needs some reworking. Maybe before we worry about specific wording, we should try to gauge agreement on some possible principles.
 * Also, possibly some of the prior disagreement was based on an abundance of numbers in body text. Maybe that should be addressed. Maurreen 16:27, 12 July 2005 (UTC)
 * Ideally there should be a technical solution to the problem, such that each user could specify what units they want to see and wikipedia would dynamically output the correct unit. This would address all unit problems, including format issues.
 * You would need to be able to specify quanitity, unit type, precision, auto scale to the most appropriate unit or not (3000 mm would show as 3 m), and whether to force the display of a specific unit where it is required. \quantity {35} \unit {kg} \precision {0} \noscale --Thax 17:05, 12 July 2005 (UTC)
 * A technical solution will not be available. See r3m0t talk 19:54, July 17, 2005 (UTC)

Possible principles for units and systems of measurement
Please feel free to add to this list. Maurreen 16:27, 12 July 2005 (UTC)


 * 1) Broad clarity to speakers of English is a priority. Corrolaries:
 * 2) Giving conversions to other systems is generally encouraged.
 * 3) An exception is casual usage such as in "Their views were miles apart."
 * 4) In general, there is no preference among metric, Imperial or customary U.S. units.
 * 5) Exceptions are made for articles that focus specifically on a region or field. In those cases, measurement units typical to that region or field are preferred. (Please note that, concerning acres and the United States, this would contradict the current style for land area.)

Suggested rewrite of 'units' section in Manual of Style
My suggestion is to rewrite the entire 'units section. Here is a suggested replacement:


 * Wikipedia articles are intended for people anywhere in the world. Try to make articles simple to read and translate.
 * Metric units are permitted. This applies to all parts of all articles.
 * The sequence of units indicates the source value. Show the source value first.
 * Show secondary units in parentheses.
 * Show the secondary units in symbolic form.
 * For example '100 mm (4 in) pipe for 10 miles (16 km)'. This that the source values are 100 mm and 10 miles.
 * Converted values should use a similar level of precision as the source value. For example, "the Moon is 400,000 km (250,000 miles) from Earth", not "(248,548.477 miles)".
 * Use standard symbols: metre -> 'm', kilogram -> 'kg', inch -> 'in' (not " or &Prime;), foot -> 'ft' (not ' or &prime;)
 * Do not append an s for plurals of unit abbreviations. Thus 'kg', 'in', 'yd', 'lb' not 'kgs', 'ins', 'yds', 'lbs'.
 * Use a forward slash for division. Thus ft/s rather than fps.
 * Non-metric units often have several versions. Be specific. Thus 'US gallon' or 'imperial gallon' rather than just 'gallon'. This is particularly important in nautical and aviation articles where 'mile' is ambiguous and should be 'nautical mile' or 'statute mile'.
 * The reader should see a space between the value and the unit symbol: "25 kg", not "25kg".

Bobblewik (talk) 19:27, 12 July 2005 (UTC)
 * I disagree with « Use a forward slash for division. Thus ft/s rather than fps. ». The writing rules for SI symbols are well known, but they do not apply to the U.S. and Imperial systems. The mile per hour could be written mi/h but the traditional English abbreviation is mph, and that is what should be used. The Conversion of units table goes to great lengths to document these traditional abbreviations.
 * Talking about ambiguities, the statute mile (a.k.a. the U.S. Survey mile) is not the same as the "standard" mile. See Conversion of units.
 * Finally, the rule of space between number and symbol does not apply to the degree of temperature (°C, °F, and others) nor to the angular degree, minute and second of arc.
 * Urhixidur 03:38, 2005 July 13 (UTC)
 * ISO 31-0 and all other formal standards on SI units require a space in the temperature 20 °C, but none in the angle 90°! Markus Kuhn 16:02, 14 July 2005 (UTC)

There may be a point of contention here, since British style calls for a space between the number and the degree sign. Oxford Style Manual (2003), 7.5: "There is a space of the line (or thin space in scientific and technical work) between the figure and degree sign. . . When giving a range of temperatures, print the symbol close up to the abbreviation: 10–15&deg;C." American style differs here, too, calling for closed up symbols to be repeated when giving a range, e.g., 10°C–15°C, 10%–15%. &mdash;Wayward 05:47, July 13, 2005 (UTC)
 * A "statute mile" is any 5,280 ft mile. In other words, it includes both the old U.S. definition which survives as the U.S. Survey mile, and the current international definition which is 2 parts per million shorter, at 1609.344 m.
 * In temperature in degrees on some scale, there is some split in authority about whether or not to use a space between the degree symbol and the number (But not exactly the U.K./U.S. distinction Wayward makes—NIST, for example, prescribes the space for degrees of temperature, unlike the spaceless degree signs for angles). As Wayward points out, one thing is clearer is that under the modern rules the degree symbol butts up to the symbol identifying the scale rather than to the number, and that something like "15° to 30°C" or "15° to 30° C" are both improper; don't use the degree symbol separate from the C or whatever.  For temperatures in kelvins, there is no difference from any other units. Gene Nygaard 12:12, 13 July 2005 (UTC)


 * Looks fine, but after
 * Use a forward slash for division. Thus ft/s rather than fps.
 * I would put something like:
 * But in scientific contexts SI notation is appropriate, for example m&thinsp;s?2.
 * Gdr 23:43, 12 July 2005 (UTC)
 * Is that the standard way to do it? m&middot;s?2 looks better to me. - Omegatron 23:56, July 12, 2005 (UTC)
 * That would be fine too. See SI. Gdr 00:03, 13 July 2005 (UTC)

About metric units: "Metric units are permitted. This applies to all parts of all articles." I agree with permitting metric units, but to single out metric units like this could be read as favoring metric units. Is that your intention? Maurreen 03:08, 13 July 2005 (UTC)
 * It certainly would be mine. There is no other system in scientific and engineering contexts, and even the U.S. and Imperial systems are defined in metric terms nowadays.
 * Urhixidur 03:32, 2005 July 13 (UTC)
 * I agree with all comments above.
 * I think perhaps the line about source value could be written "The first value stated should be the source value." This gets to the point in the first sentence.
 * And maybe in non-scientific contexts, metric and US equivalents should be encouraged? Saying they're "permitted" is a bit vague, though I realise that some contexts like to be different and so won't want equivalents - but apart from scientific articles, I can't think of any.  More general knowledge (as opposed to specialised) articles will probably want equivalents.
 * "Use standard symbols" - does this mean that the entire word is to be avoided (in the source value), or just don't use non-standard symbols if you want to use symbols? Neonumbers 04:29, 13 July 2005 (UTC)
 * How about "Both metric and U.S. units are acceptable" instead of: "Metric units are permitted. This applies to all parts of all articles."
 * And "Avoid nonstandard symbols" instead of "Use standard symbols." Maurreen 05:34, 13 July 2005 (UTC)

There is no inconsistency in the MoS as it stands. The MoS says the following (and these points have been added to or tweaked by several editors, and appear to represent the current consensus): There is no consistency there whatsoever. What the MoS says is that editors may choose what to do in all non-science articles. Leaving the choice to editors on the page is a good thing. Forcing them to do one thing or the other is not, and will make the MoS unpopular. SlimVirgin (talk) 05:53, July 13, 2005 (UTC)
 * 1) In scientic contexts, use SI units.
 * 2) There is no need to perform a conversion.
 * 3) In non-science articles, editors may choose metric, U.S. customary units, or Imperial.
 * 4) Conversions from and to any of the above may be helpful, but are not mandated.
 * 5) Specifically, conversions should not be offered if they interfere with the quality of the writing.
 * 6) Multiple equivalents are cumbersome and should be avoided, except for unusual units.


 * I've removed my personal comments about Bobblewik's motives, and I apologize for having made them. I accept that he's is acting in good faith. It's just that I find this issue a little frustrating, but that's no excuse for the ad hominem remarks. I'll also apologize to Bobblewik on his talk page. SlimVirgin (talk)
 * Tsk, tsk. No one is trying to force you to do one thing or the other. You are trying to prohibit the addition of valuable information in articles that you edit. In addition, you have repeatedly tried to cast this as a disagreement between reasonable editors and one maniacal single-issue editor, ignoring the fact the many people here have supported the addition of metric units. This is much more about your stubbornness than about Bobblewik's. Rl 09:27, 13 July 2005 (UTC)
 * There is inconsistency - to say that they both should be given and there is no need for them is inconsistency. And to say on a manual of style that editors can choose which style they prefer defeats the purpose of having a manual of style.  Even if the principles are consistent, the text is not and should be changed.  Bobblewik is not alone in his view.  He is not misrepresenting the situation.  People come here for a guide as to what to do.  To say "do what you want" is to say "it's your choice" to someone who asks for advice.
 * With metric/U.S. units, I think that both should be stated, first the source then the other version in brackets, except in scientific articles and other contexts which I can't think of. In most contexts both should be there (and thus mandated as far as the MoS can mandate things), so (if people agree with me on this point) that should be explicitly stated.  Neonumbers 09:56, 13 July 2005 (UTC)
 * Symbols for units of measurement should never be italic (as Bobblewik has done them).
 * The polar exploration context is another one in which miles should always be identified. This Google search dealing with Ernest Shackleton's closest approach to the South Pole shows a far too common conversion error (an error which Wikipedia used to share with many other sites) resulting from a failure to identify miles. That problem results from failure to identify the miles as nautical.  Of course, since nautical miles are the conventional units in this context, if statute miles are used instead they should be specifically identified as well. Gene Nygaard 12:34, 13 July 2005 (UTC)
 * I prefer the existing text, as quoted by SlimVirgin. Especially number 6 - we often deal with the problem of odd source units in automobile articles.  Two specific and common examples are PS and cubic inches in engine displacement.  In both cases, I like to provide the more common equivalents (kW and hp for the former and cc and L for the latter) in articles.  We simply must leave the original unit in the text, since it's historically significant.  But since they're also unfamiliar, and since automobiles are universally enthused upon, it is courteous and helpful to provide more familiar units as conversions.
 * I would add one more rule: If conversions of metrics are available from a source, these shall be used rather than a more "accurate" recalculation.  In many cases, a manufacturer will intentionally round up or down, or even recalculate a value for two different markets.  This happens often with the kW/hp measurements.  In these cases, we must use the manufacturer-provided metrics.  Bobblewik pointed this out to me, and he is entirely correct.  --SFoskett 13:14, July 13, 2005 (UTC)

Unchallenged changes and some of my rewording thus far make the rewrite (changes to original in non-italics, original in italics. It will not be in italics in the manual, so ignore the units being in italics here):
 * Wikipedia articles are intended for people anywhere in the world. Try to make articles simple to read and translate.
 * Metric units are permitted. This applies to all parts of all articles.
 * Always use the source value (e.g. from manufacurer's specifications). Write this value first, then converted units.
 * Show converted value units in parentheses.
 * Show the converted value units in symbolic form.
 * For example '100 mm (4 in) pipe for 10 miles (16 km)'. This that the source values are 100 mm and 10 miles.
 * If a source already has a converted value, use that rather than calculating the conversion yourself.
 * Converted values should use a similar level of precision as the source value. For example, "the Moon is 400,000 km (250,000 miles) from Earth", not "(248,548.477 miles)".
 * Use standard symbols when using symbols: metre is 'm', kilogram is 'kg', inch is 'in' (not " or &Prime;), foot is 'ft' (not ' or &prime;)
 * Do not append an s for plurals of unit abbreviations. Thus 'kg', 'in', 'yd', 'lb' not 'kgs', 'ins', 'yds', 'lbs'.
 * For division (as in kilometres per hour):
 * In scientific contexts, use scientific notation and with thin spaces (&amp;thinsp;), for example kg&thinsp;m&thinsp;s-2, g&thinsp;mol-1
 * In non-scientific contexts with metric units, use a forward slash, for example km/h
 * With non-metric units, use the standard abbreviation, for example mph, fps
 * Non-metric units often have several versions. Be specific. Thus 'US gallon' or 'imperial gallon' rather than just 'gallon'. This is particularly important in nautical and aviation articles where 'mile' is ambiguous and should be 'nautical mile' or 'statute mile'.
 * The reader should see a space between the value and the unit symbol, for example "25 kg" not "25kg".
 * But with temperatures in degrees Celsius/Fahrenheit and geographical co-ordinates, there is no space, for example "25°C" not "25 °C". Note that Kelvin does not use the degrees symbol.

I didn't change the part about Metric/U.S. units to be allowed/permitted. The part about flow of sentences (e.g. "Legend has it that the hero was a thousand miles from the volcano when it erupted." (fictional)) is implied, in my opinion, but that can be explicitly said if it needs to be. Neonumbers 02:26, 14 July 2005 (UTC)
 * Many sources make bad conversions, in any direction. To say that we should use bad conversions just because our sources do is sheer stupidity.  What you may sometimes have is an interpretation question, in determinining which of the values given in the source is most reliable and accurate. Not everybody follows the rule of putting the original first, for example, and there are cases where the measurements can be equally valid in either measurement given, even though conversion from one to the other in either direction wouldn't give the same result (this usually happens when the value is actually known to more precision than what is stated in the article). Sometimes, where it is more of a naming question than a measurement question, or a case of making the measurements in slightly different ways in differnet units, there can be discrepancies (the AR-15 and M-16, with ammunition designated by either of the names .223 Rem or 5.56x45 mm NATO, for example, where 0.223 inch is not equivalent to 5.56 mm).  But unless you can tie it to different testing conditions on different standards, and specify those conditions in some way, I don't see how that would apply to automotive engine power questions.  In that case, it is probably most often an interpretation question involving the namufacturer's meaning of "horsepower", and whether that unit is 75 m·kgf/s or 550 ft·lbf/s. Gene Nygaard 14:34, 14 July 2005 (UTC)
 * Another naming example is floppy diskettes, where there is no need to express the conversion more precisely than 3½ inches (for example, 3.543 in), unless the actual standards for their manufacture are being discussed. The weird thing is that the manufacturers did such a good job of pulling the wool over our eyes that they even fooled their advertising people at times, with some manufacturers labeling these diskettes as "3½ in (88.9 mm)" even though they are actually designed to be 90.0 mm.  This is also a good, specific example of a case in which we should not follow the manufacturer's bad conversions.  Gene Nygaard 14:43, 14 July 2005 (UTC)
 * Does anyone know the syntax for the thin space? In my browser (IE6), &amp;thinsp; is a really wide space.  Neonumbers 12:36, 14 July 2005 (UTC)

I suggest that knots are used for speed at sea, and only knots - no km/h. As far as I know no seaman use anything but knots. This is also related to the grid system used for navigation - a knot is one nautical mile, pr hour. So I propose that the use of km/h in parenthesis be dropped. Ulflarsen 13:29, 14 July 2005 (UTC)
 * I disagree. We are not (only) writing for seaman. Rl 14:04, 14 July 2005 (UTC)
 * We aren't writing just for seamen. They're a miniscule part of our audience.  One important thing about SI is that it is interdisciplinary as well as international.
 * When I was in the U.S. Army many years ago, our maps had tick marks for latitude and longitude in grads or grades (another fairly recent name is gons). Of course, 1 centigrad(e) is to 1 kilogmeter as 1 minute of arc is to one nautical mile.  Gene Nygaard 14:18, 14 July 2005 (UTC)

This discussion is hard to follow
The discussion above is getting complicated, in my view. I think it's too broad and would like to suggest it be broken down. Maurreen 06:22, 16 July 2005 (UTC)

Acceptability of both metric and customary U.S. units
My suggestion is for the style guide to include something like this: "Generally speaking, both metric and customary units are acceptable, and conversions are encouraged." Comments? Maurreen 06:22, 16 July 2005 (UTC)


 * Seems sensible to me, although "customary" is ambiguous. Maybe it's best left ambiguous, as it could mean US imperial, British imperial, traditional Indian units, etc. "Customary" doesn't quite seem to be the right word, though I admit I can't think of a better one, jguk 18:41, 16 July 2005 (UTC)
 * I left a word out. I meant to say either metric or customary U.S. units. I agree that the phrasing may be imperfect, but I can't think of anything better either. Maurreen 19:18, 16 July 2005 (UTC)
 * Well, I will complain if "U.S." is put in there, but not if it is left out. Gene Nygaard 16:03, 21 July 2005 (UTC)


 * This is a style guide! It lists best or recommended style. (More a goal than a rule.) Using US/English-only measures is surely not best style (ambiguous, widely unknown), whereas it is perfectly acceptable to have metric units without any conversion given at all.
 * If the source/legal value is English/Imperial/US customary, fine, use it and give the metric equivalent in parenthesis. If the source value is in any other non-metric unit, use that and also give the metric equivalent, but no other. Otherwise use the metric value only.
 * “SI units are recommended. If using non-metric units, give conversions in parenthesis.”
 * I can’t stand this wishy-washy, which also applies to other areas. Christoph Päper
 * To Christoph Päper, I never said "only". Maurreen 15:09, 18 July 2005 (UTC)
 * Your wording would allow US only, because it allows US units in general and only encourages, but not requires, conversions. We certainly shouldn't allow any other customary system, except when used by the source. Christoph Päper 00:56, 19 July 2005 (UTC)

We really shouldn't add "US". Why should we favour US units explicitly? What's wrong with traditional British or Indian units, say? I think explicitly encouraging US units, but remaining silent on all other well-known and familiar units (depending on which part of the world you're in) is a recipe for arguments and divisiveness. Perhaps if we replace "customary" with "traditional" (and leave out any national determinants) it'd look better, jguk 19:48, 16 July 2005 (UTC)
 * Are traditional British units still commonly used? My understanding is that the only country remaining of the English-speaking ones to hold out with a standard non-metric system is the U.S..  I dare say Indian units wouldn't really fit on an English encyclopedia, except in articles that are about something Indian.  Now, I'm not saying I think the U.S. should keep its non-metric ways.  But the fact is that they do - and to my knowledge are the only ones to do so.  Even if people in say, Britain are familiar with the traditional ones, they'll also know the metric ones, won't they?
 * My fear is that saying any non-metric or traditional unit will lead to the abuse of it - the listing of endless units. Sure, no-one will be stupid enough to do that but we don't know that.  In most contexts, I cannot think of any necessary units except for metric and U.S..  Obviously some specialised articles, e.g. those concerning speed at sea (knots?), will need otherwise.  But perhaps, just once, I'll want to favour the United States with this (nothing against Americans :-P).
 * Correct me if my understanding is wrong. Neonumbers 12:45, 18 July 2005 (UTC)
 * Neonumbers put it well. Maurreen 15:09, 18 July 2005 (UTC)
 * In the U.K., highway signs cannot be in kilometres. Draft (draught) beer may not be sold by the litre.
 * Even though the troy pound was outlawed in Great Britain in the 19th century, today the troy ounce still enjoys a specific exception from the metrication laws of the U.K. and of Australia and probably others.
 * In Canadian livestock auction markets, they sell cattle in dollars per hundredweight (100 pounds or 45.36 kg, of course, unlike those people who think hundred is written in digits as "112"), and hogs in dollars per hundred kilograms.
 * Need more? Look at the aviation industry, for example.  Shipping.  Or bicycle spoke tensions which are often "metric" even in the United States, but not SI.  Same for food calories. Gene Nygaard 16:14, 21 July 2005 (UTC)


 * But in all these cases, we'd be looking at the source value anyway &mdash; its use mandated because of its being the source.
 * Unless, for the most part, the British (for instance) know only of British units and stumble into confusion when they see "200 metres" or "30 litres" in much the same way I would when I see "15 feet" or "7 ounces", I can't see why we should be using anything aside from metric and U.S. units as converted units.
 * If there is an English-speaking country other than the U.S. not familiar with the metric system, then by all means, include that too. But unless you're telling me that British and Canada are tied to their own, independant, traditional systems in much the same way the Americans are, there is no reason that I can see to specifically convert a metric unit into these for inclusion within parentheses.
 * Perhaps this needs clarity in wording: always use the source value and unit, and convert so that metric and U.S. units are both stated as either source or converted, thus at the most we would have three units. We definitely shouldn't be converting anything into a unit of only a non-English speaking country.
 * (Of course, if the U.S. have come to grips with the rest of the world, then we won't need U.S. units either.) Neonumbers 06:27, 23 July 2005 (UTC)

My biggest problem lies with horsepower. The hp unit is not similar in the US and in the Continental Europe, where "hp" is an english literal translation of the German PS. The PS unit is also used in Japan. In automobile articles, the "hp" unit is used as a de facto unit, but it's no longer lawful in most of the world. Most articles have a "## hp (## kW)" format, but this is somewhat error inducing, as some people might convert the kW from the hp and come to a wrong value for European market and Japanese market cars. It's been suggested that PS should be converted to hp, but I am against it, as the hp is not used in Europe and Japan. As a compromise, I've been using "## PS (## hp/## kW). --Pc13 15:38, July 21, 2005 (UTC)

Third attempt at a revised wording for units section
It now specifies that non-metric units are permitted. It no longer addresses editorial problems that are interesting but do not appear to cause holy wars.
 * Wikipedia articles are intended for people anywhere in the world. Try to make articles simple to read and translate.
 * Metric units are permitted. This applies to all parts of all articles.
 * Non-metric units are permitted. This applies to all parts of all articles.
 * Try to get the best source value (e.g. from manufacturer's specifications). Write the source value first and then the converted value. Use digits and unit symbols for the converted value. For example, '100 millimetre (4 in) pipe for 10 miles (16 km)'. This means that the source values are 100 millimetres and 10 miles.
 * Converted values should use a similar level of precision as the source value. For example, "the Moon is 400,000 km (250,000 miles) from Earth", not "(248,548.477 miles)".
 * Use standard symbols when using symbols: metre is 'm', kilogram is 'kg', inch is 'in' (not " or &Prime;), foot is 'ft' (not ' or &prime;)
 * Do not append an s for plurals of unit abbreviations. Thus 'kg', 'in', 'yd', 'lb' not 'kgs', 'ins', 'yds', 'lbs'.
 * Non-metric units often have more than one version. Be specific. Thus 'US gallon' or 'imperial gallon' rather than just 'gallon'. Similarly, 'nautical mile' or 'statute mile' rather than just 'mile' in aviation, space, sea and some other contexts.
 * The reader should see a space between the value and the unit symbol, for example "25 kg" not "25kg".


 * Works for me. Thank you. Maurreen 13:11, 16 July 2005 (UTC)
 * Does anyone reading this have an idea of how hard it would be to patch the wikipedia software so that it is aware of units? The edit text could for example look like this: Measure:100 mm, and be presented like "100 mm (4 in)" or "100 mm" or "4 in" etc, depending the user's preferences. Just an idea, anyway; I don't have time to look into it myself at the moment. Grahn 19:08, 16 July 2005 (UTC)
 * I think the part about division and the part about the degrees symbol should be there. I don't see who could start a revert war on those issues (really really strange people don't count).
 * The degrees symbol part is especially important because otherwise people might think a space is meant to be there as well.
 * The part about division is especially importnat for science articles - for which it is imperative to have a standard style (is there a MoS for science articles?), but I think it's a good idea to include that for non-science articles too. Neonumbers 01:49, 18 July 2005 (UTC)
 * A space should be there before a degrees of temperature symbol (37 °C), according to several authorities. E.g., NIST Guide for the Use of the International System of Units, section 7.2:  "The symbol °C for the degree Celsius is preceded by a space when one expresses the values of Celsius temperatures".  Also,  National Physical Laboratory where the general rule doesn't mention any exception for degrees, and the NPL practice includes those spaces.
 * There is no good reason to specify exponent notation for units formed by division in a scientific context. It would be a much more useful guideline to make sure that editors know that they cannot have two divisions in a unit, without disambiguating it with either parentheses or that exponent notation.  Gene Nygaard 08:51, 18 July 2005 (UTC)
 * There is no ambiguity in foo/far/baz! It is always equivalent to foo/(bar·baz) == (foo/bar)/baz == (foo/baz)/bar.
 * There is ambiguity in foo/bar·baz only&mdash;actually there shouldn't be one either, because it means (foo/bar)·baz == foo·baz/bar by all math everyone learned in school, but people (incl. scientists) are lazy and/or stupid, taking it to mean foo/(bar·baz) in some cases. Christoph Päper 01:08, 19 July 2005 (UTC)
 * IUPAC Compendium on Chemical Terminology (1997) also uses spaces in 25 °C, etc. small pdf file. Gene Nygaard 09:02, 18 July 2005 (UTC)
 * The BIPM (the authority that maintains SI) uses spaces in 25 °C etc. Markus Kuhn reports that ISO 31-0 and all other formal standards on SI units require a space in the temperature 20 °C. I am not aware of any rational reason or good authority for making temperature an exception to the simple rule.
 * As far as division is concerned, a single forward slash is indeed legitimate in science. There is even a slash in the definition of the metre, which is about as scientific as you can get. Ambiguity occurs when people use two or more slashes for division. But there is no ambiguity with one.
 * We could adopt conventions opposite to those of the BIPM and ISO 31. But I don't know of a good reason. Bobblewik 10:35, 18 July 2005 (UTC)


 * Okay... didn't know that degrees Celsius is meant to have a space - someone said it somewhere up there and I assumed they were correct. So, space with all units.  Makes sense.
 * Because the page Bobblewik referred to didn't really have any units with division on it (just numbers, and numbers in fractions - well, they should use forward slashes), that's not really somewhere to point to for this purpose - it shows nothing about unit conventions.
 * But I did do some looking up, and the SI brochure on that same site does say that either slash or indices may be used, and that with ambiguity and slashes (which can also occur in e.g. m·kg/s³·A), use brackets (m·kg/(s³·A)). So convention is either or.  Makes sense too.
 * Given this, it does make much more sense to standardise the solidus in the manual (a quick jump around the science articles showed this in place), and to ask that ambiguity be solved with brackets not exponents.
 * That brings me to, should we specify that ambiguity be solved with brackets (which means, no negative exponents - I've yet to find some in Wikipedia)? For consistency, I think this is a good idea.  I saw this used on that box that appears down the right hand side of the chemical elements.  On the other hand, the prescribing of slashes already implies the use of brackets in ambiguous division, and it's already a rare case anyway.  It also implies not using negative exponents.  Anyway, I'm happy with specifying solidi.  Neonumbers 12:23, 18 July 2005 (UTC)
 * There are a number of Wikipedia articles which do use negative exponents. I see no reason to specify one way or the other.  I prefer a solidus if only two units are involved; you don't run into any ambiguity problem there.  But if most of them are more complex units justifiably expressed in unambiguous negative exponents, then making the whole article consistent makes sense.  If it becomes an issue in some particular article, the editors can hammer it out.
 * The articles about chemical elements generally use units written as "W/(m·K)" rather than W·m&minus;1·K&minus;1, but in general I'd find either acceptable. You could get some  consistency in removing all the old "W/(m*K)" or "W/m*K" (the multiplication after a division sign, without grouping, also causes some ambiguity problems), and specific projects such WikiProject Elements as will achieve additional consistency within a more limited scope.
 * As far as the degree sign goes, when it is used for angles it is still written closed up to the number, as are the apostrophe/prime signs for minutes and seconds. There was a discussion in regard to latitude and longitude in one of the projects (I think it started on the talk page here, then on the geographical coordinates project), as to whether a space or thin space or no space should be used between the degrees and the minutes and the seconds, with no clearly applicable  general rule and much of the discussion dealing specifically with how they look in Wikipedia implementation.  Gene Nygaard 13:39, 18 July 2005 (UTC)
 * I still think, for the pure and sole sake of consistency, we should specify one or the other - slashes because and only because they're already more in more widespread use here. I wouldnt' be surprised if someone, writing a science article, wanted a guide as to which to use, because after all, both are standard.  Because it there's no difference with science articles, I don't see a point in specifying brackets or no negative exponents or anythign specifically for science - just something like, "For division with metric units, use slashes, e.g. km/h; and with traditional units, use the traditional abbreviation, e.g. mph.", and that would apply to all articles.
 * So I'm not concerned about ambiguity anymore - it is a very rare case. But division shoudl be touched on - using solidi for metric, appropriate abbreviations for non-metric.  Neonumbers 10:51, 19 July 2005 (UTC)

After the third bullet point I would add a new one: "Converted values should ordinarily be given in parentheses, but if conversion is inserted into a direct quotation, then brackets should be used." This follows the general rule that a direct quotation should not be altered unless the alteration is indicated, with brackets for insertions and ellipses for omissions. Also, I trust that some common sense would apply to forestall excessive conversion. If an article happens to quote "A journey of a thousand miles begins with a single step," surely the thought is clear to our metric friends without the insertion of a kilometer value to ruin the cadence. JamesMLane 13:24, 22 July 2005 (UTC)

Measurement systems No. 4
How about this: "Generally, metric, Imperial, or U.S. customary units are equally acceptable, and conversions are encouraged." Maurreen 06:54, 23 July 2005 (UTC)
 * What about Indian customary units (particularly on Indian subject matters)? We are an international encyclopaediaia after all. Also, particularly for precious metals and jewellery, shouldn't we allow troy weights and carats? jguk 07:12, 23 July 2005 (UTC)
 * Briefly --
 * My proposal does not disallow anything.
 * Much of life, style guides in general, and my proposal in particular, are not intended to cover every single situation that might ever arise. For one thing, this is indicated by the word "generally" prefacing my proposal. Maurreen 07:22, 23 July 2005 (UTC)
 * What about "Where giving measurements it is often beneficial to provide conversions to help readers unfamiliar with one style. For example, you may write 8kms (5 miles) rather than just 8kms"? jguk 07:39, 23 July 2005 (UTC)
 * Shouldn't this be: write "8 km (5 miles) rather than just 8 km"? &minus;Woodstone 09:24, July 23, 2005 (UTC)
 * That should be "8 kilometres (5 miles)" or "8 km (5 mi)" or "8 kilometres (5 mi)".  You can't spell out the converted and abbreviate the source. Neonumbers 11:30, 23 July 2005 (UTC)
 * It probably should rather read “eight kilometres (five miles)”, “eight kilometers (five miles)” or “8 km (5 mi)”, or, if I was the dictator of Wikipedia, “eight kilometres” or “8 km”&mdash;one system of units and one set of spelling rules should be preferred (not required). Christoph Päper 14:59, 23 July 2005 (UTC)
 * Would it not be best to use always abbreviations within parentheses? For example, from Manual of Style (dates and numbers):
 * ". . . 'the Moon is 400,000 km (250,000 miles) from Earth,' not '(248,548.477 miles).' "
 * It should read: ". . . 'the Moon is 400,000 km (250,000 mi) from Earth,' not '(248,548.477 mi).' " Or, ". . . 'the Moon is 400,000 kilometers (250,000 mi) from Earth,' not '(248,548.477 mi).' " &mdash;Wayward 11:52, July 23, 2005 (UTC)


 * It should be "8 kilometres (5 miles)" or "8 km (5 mi)"; the spelled out version could be italicized, but even if the test around the latter is italicized, it should then be "8 km (5 mi). But also, while there are rules about not mixing spelled out words and symbols within one unit (e.g., "miles/s"), there are times when using conversions with one spelled out and the other using symbols makes sense:  for example, if the spelled out unit or even just its symbol is one likely to be unfamiliar to many readers.  Furthermore, there is no good reason for generally including a non-breaking space between a number and a unit.  Sometimes it is helpful, but it should not be our general rule, expecially since including all that &amp;nbsp; makes it more difficult for editors to read the edit screen.   Gene Nygaard 14:20, 23 July 2005 (UTC)

I can just imagine encountering in the basketball article: "An international basketball court is 28 metres (92 feet, 61 hasta, 24 pou) wide by 15 metres (49 feet, 33 hasta, 13 pou) long." Hmmm. Yeah. (I had to do a quite a bit of research to produce those figures. They are, btw, metric, imperial/U.S., Indian and Chinese in that order.) Units in specialised articles will always be a bit of a troublemaker, with interpretations and the like. I get the impression specialised articles are a lot more common than I thought they would be. So, how about:
 * The source value should be in whatever unit is appropriate for the context. You should also help readers by converting these to metric and customary U.S. units, whichever it is not or both if they should be required.

Perhaps the wording could be better. My point is, people are advocating specialised articles &mdash; say, Indian or diamonds or sea distances or selling something by the hundredweight &mdash; and clearly the units they state are appropriate as the source. But with converted units, there is never a need to go past metric and U.S. (I would think). (There would be some special cases where either or both sets of converted units are done away with due to context, because no-one would ever think to use them there... dunno if that's worth mentioning though.) Neonumbers 11:30, 23 July 2005 (UTC)
 * According to Chinese units 28 × 15 m² would be 16.8 × 9 步² (bu, pou …) or 8.4 × 4.5 市丈² (zhang). We should not require more than one conversion, metric being preferred of course. There are some special cases, though, i.e. where a custom unit had been (re)defined in terms of English units, that not only happened in the British Empire, but also for instance in Czaristic Russia.
 * IMHO there are only these acceptable forms in order of preference:
 * source/legal (metric conversion to equal precision),
 * metric source/legal,
 * metric source/legal (US customary conversion to equal precision),
 * metric,
 * metric (US customary conversion to equal precision).
 * Christoph Päper 14:59, 23 July 2005 (UTC)

If we're to maintain friendly diplomatic relations, we're best off not referring to the US as a special case. I can foresee arguments between UK and US gallons now. And of course, taken literally, it would mean that a British article on beer could not refer to it being sold in pints. To comply with the proposed guideline it would have to say 0.568 litres (0.9 pints) (with the pint being a US pint!). Why do we need to say anything specific anyway - surely the general principles of commonsense and thinking of the reader will always yield a sensible solution? jguk 12:03, 23 July 2005 (UTC)


 * There is the “imperial” prefix (maybe capitalised) to denote that pint as the legal unit. Christoph Päper 14:59, 23 July 2005 (UTC)


 * The above proposal in italics seems eminently wise to me. It might be indicated that there is no burden on the fist editor to do the conversion. The other units should just be respected if they are added by other editors. This leads to the following refined proposal:
 * The primary value should be in whatever unit is appropriate for the context, especially if it is the source value. The original or other editors are encouraged to help readers by adding conversions to metric and traditional English language units, whichever is (are) not present yet. These conversions should be respected by all subsequent editors.


 * A history of pointless revert wars has shown that it would be highly preferred to have a firm guideline. &minus;Woodstone 12:13, July 23, 2005 (UTC)


 * (edit conflict, adding anyway) Sadly, we can't rely on commonsense here. The reason this discussion started almost two weeks ago was that some editors would insist that there only be customary U.S. units in "their" article. And the initial debate was not whether to recommend adding both units, but whether it is okay to remove metric units where other editors added them. Rl 12:18, 23 July 2005 (UTC)


 * As RI says, the proposal to reword the the Manual of Style started because some non-metric editors remove metric units. Other editors want metric units. The test question in that conflict was:
 * Does the Manual of Style permit removal of metric units?
 * The claimed answer was Yes on the basis of their interpretation of subjective elements of the wording. Any revised wording will face that test question. That was why the word permitted was used instead of generally and recommended.


 * As far as other issues are concerned, the Manual of Style is particularly useful when:
 * (a) there is confusion that causes a clear and present problem
 * or (b) there is conflict that causes a clear and present problem
 * There are plenty of good ideas and interesting nuances of wording, but shall we just try to solve the problem of prohibition of metric units? Bobblewik 12:44, 23 July 2005 (UTC)


 * That is precisely what above formulations are aiming at. Only a well balanced formulation, doing justice to the many viewpoints can hope to be included as the guideline. The above formulation does not mention any particular country. The mention of the language in which this wiki is written should not be objectionable. It allows other units than the disputed ones in specialistic articles (aeronautics, india). It does not prefer any of the disputed units. It does not open the door to endless lists of conversions. It does not force anyone to include conversions. It just might convince former hard-liners to be lenient, because of its evenhandedness. &minus;Woodstone 14:50, July 23, 2005 (UTC)
 * I see Woodstone's proposal as very sensible. Maurreen 17:35, 23 July 2005 (UTC)


 * With beer and British pints, by my proposal, the "correct" form would be x pints (x litres, x U.S. pints). I have no intention to ban British units.  I also have no intention to mandate U.S. conversions - only metric ones.
 * btw, I'm sorry about those Chinese unit conversions with that basketball court... I had an external source guide me because there is no article on Indian units that I could find here. lol.
 * (this comment continues as the second comment in the next section) Neonumbers 11:28, 24 July 2005 (UTC)

Units and style guide deletions for consistency
If we use the proposal above ...
 * The primary value should be in whatever unit is appropriate for the context, especially if it is the source value. The original or other editors are encouraged to help readers by adding conversions to metric and traditional English language units, whichever is (are) not present yet. These conversions should be respected by all subsequent editors.
 * ... is there anything we should consider removing from the style guide to make it consistent with the proposal? Maurreen 17:35, 23 July 2005 (UTC)
 * (continued from previous section): My only concern is that I wonder if conversions and only conversions to traditional British units should ever be necessary to assist a reader who would not know otherwise. If they are, please just say so &mdash; I don't know how Britain is with the metric system &mdash; so that I can stop going on about how only the U.S. is still attached to their own non-metric ways.  For all I know, Britain might be.  And Canada.  And Australia.  New Zealand's not, I can tell you that much ;-)
 * And if the U.S. is already metricised, then say that as well.
 * But I still don't see a reason to include conversions and specifically conversions for any system apart from those commonly used in the English language. These, by my understanding, are only metric and U.S..  If you don't want to say U.S., that's fine by me, if you can think of a better way to put it.  I'm not too confident about common-sense restricting it to these.
 * Please inform me about how metricised Britain and other English-speaking countries are. Aside from that, the above wording's good.
 * (To my knowledge that wording would not create inconsistencies.) Neonumbers 11:28, 24 July 2005 (UTC)

- This discussion continues on archive 24 >>