Jump to content

Wikipedia talk:Manual of Style/Dates and numbers/Archive 130

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 125Archive 128Archive 129Archive 130Archive 131Archive 132Archive 135

Sharon Johnston date format

Talk:Sharon Johnston lists the issues. Feel free to add to the discussion there. --Walter Görlitz (talk) 16:53, 7 August 2010 (UTC)

Permitting metric distances in road tables

Wikipedia lists distances in American articles about roads. I added metric units but was reverted. When I queried this, I was told that it had been discussed in the a subsection of the manual of style:

There is nothing in the MOS that forbids it but I was told it was the de facto standard following those discussions. I read the discussions and still think that it should be permitted to use metric units for distances in a table. I can understand that there might be something special about "milepost 17" but I can't see what is special about "the second exit is 1.8 miles from the origin and the third exit is 2.4 miles from the origin". What do other people think? Lightmouse (talk) 21:39, 1 August 2010 (UTC)

Quoting MOS:CONVERSIONS: "When units are part of the subject of a topic—nautical miles in articles about the history of nautical law, SI units in scientific articles, yards in articles about American football—it can be excessive to provide conversions every time a unit occurs. It could be best to note that this topic will use the units (possibly giving the conversion factor to another familiar unit in a parenthetical note or a footnote), and link the first occurrence of each unit but not give a conversion every time it occurs." This has been used in the past to allow road junction lists to use only the appropriate measurement unit in the table. Other measurements are converted, especially the total length of the road in the infobox and prose. For some countries and situations it is appropriate to list both units. In the UK, the driver location signs, which are the basis for the distances in the table, are denoted in kilometres but the odometers and other signage are all in miles. Some US highways are measured in metric, like Interstate 19 and Delaware Route 1. Otherwise, it is felt that including both, especially in the manner here [1] clutters the table too much. Imzadi 1979  21:49, 1 August 2010 (UTC)

The phrase "every time it occurs" was intended to apply to repeats of the same value. This is not the case with a table of different distances. When you say that some felt that metric units clutter the table, were those people from metric countries? Lightmouse (talk) 21:56, 1 August 2010 (UTC)

There were members of CRWP and UKRD in the discussions earlier this year on overhauling MOS:RJL. The UK is using two columns for distances, the US and Canada are using one. The UK is an exception because their driver location signs, their equivalent of the US milepost signs, are in metric while their odometers and the other signs on the road use miles. The other countries use the system that's appropriate to their country's road. Imzadi 1979  22:02, 1 August 2010 (UTC)
Second thought, I don't see that the quoted passage refers to the same value repeated. It refers to repeated usages of the same unit of measurement, not the same value. Imzadi 1979  22:19, 1 August 2010 (UTC)
Ideally, I think road junction lists should include both distances in miles and kilometers, as Wikipedia is supposed to provide a worldwide view and needs to include numerical units in both English and metric. However, the junction list may look a little clunky with both distances. IMO, I wouldn't be opposed to both miles and kilometers being in junction lists. Dough4872 22:51, 1 August 2010 (UTC)

It is way too much clutter, especially for really lengthy junction lists. --Rschen7754 00:11, 2 August 2010 (UTC)

In my view the junction lists should reflect what the motorist actually sees on the signs and other markers, not only in terms of units of measure, but also in terms of towns and locailities. I, for one, have on occasions printed off the junction list for navigation purposes when preparing for a long trip. Martinvl (talk) 07:41, 2 August 2010 (UTC)

Requiring the roads projects to include both units of measure in all junction lists amounts to requiring roads editors to do a lot of extra work (some junction lists have hundreds of junctions) that would provide a very small benefit, or none at all. --Rschen7754 07:46, 2 August 2010 (UTC)

Requiring, allowing, and forbidding are three different things.
Don't do any extra work. Just stop enforcing the prohibition. Lightmouse (talk) 09:49, 2 August 2010 (UTC)
There's not really a "prohibition" per se. Rather there is a consensus that adding the addition measurements is unnecessary. Additionally, in cases where both systems are in use, they're not done in the manner you did to Capitol Loop. There are two separate columns with the units as the header. By using the {{convert}} like that, the table was extra cluttered with the parentheses and units in each cell. Forgive us if some of us are a bit defensive, but in the past it has been the practice by non-roads editors to "require" roads articles to make changes, yet expect the roads editors to make the required changes without additional help. Imzadi 1979  10:24, 2 August 2010 (UTC)

Ah, it sounds like a resolution is possible that doesn't include 'require' or 'forbid'. It's now merely a matter of preferring separate columns. Lightmouse (talk) 11:10, 2 August 2010 (UTC)

Are we recording what is actually displayed, or what is reputed to have been measured? Martinvl (talk) 11:25, 2 August 2010 (UTC)
Lightmouse, conversions were never strictly forbidden. Consensus has been not to require them, nor even use them, when a country uses one system alone. I will say that I see adding another column to tables as completely unnecessary. The total length in the infobox is converted, as are all measurements in the body of the article. That's all that's needed. When I read through a Canadian highway article that uses metric alone, the distance points in the table give me an idea of how much of the total distance has been covered up to a point. I don't need a conversion, nor would one be appropriate to me, because the signage is in metric only.
Martinvl: I can't speak to how others obtain them, but any distances given in the articles I've written are from official department of transportation sources. Nothing is measured, only transcribed from government documents. The only alteration comes with any necessary arithmetic to compensate for where the sources reset the distance to zero at an intermediate point. Imzadi 1979  11:41, 2 August 2010 (UTC)
Hi, Imzadi and others. You may remember the conversations we had when I raised the issue of the US-centric focus of the RJL MoS subpage, and whether it could be merged into the US Highways MoS subpage. I believe we managed to resolve that!

I haven't kept up with this current issue about metric conversions, but I wonder why an exception should be made for just this field. I accept that space is at a premium in tables, but other tables manage metric and US customary conversions perfectly well. The slight addition to "clutter" is just something we have to live with internationally when English-speakers have a dual system.

The Capitol Loop table seems to be ideal for adding conversions: the first two sub-columns would need very minor increases in width, and there appears to be no space crisis in the other columns, which could be narrowed a little if people want to retain the current overall width.

I am particularly concerned about the "slippery slope" implications of denying the many international readers, including many native English-speakers, easy comprehension of the distances. Tony (talk) 15:40, 2 August 2010 (UTC)

Well, Capitol Loop is not the best article for an example in this discussion. Compare that article instead to Interstate 75 in Michigan where there are around 140 rows to the table and 6 columns. There's now pressure from the geocoding people to potentially add a seventh column for coordinates. Capitol Loop doesn't have the first two columns because it's a short route in a single city and therefore in a single county. Highways with freeway (motorway) sections, or all freeway highways gain an exit number column. Some like Pennsylvania Turnpike have a column for the interchange names and even a second exit number column for old vs. new exit numbers. Once you get to the longer highways, the county and location columns pop back into the table, and it gets much wider. When I get around to fixing I-75's table to including the missing distances for the bridges, I'll update it to the full three decimal places of precision available to the measurements. As for this being US-centric, will we be forcing Canada to add miles to their tables? MOS:CONVERSIONS already says that repeatedly converting measurements is unnecessary. I think that exception is ideally suited to highway articles with the full 5–6 column complement on their tables. Imzadi 1979  16:17, 2 August 2010 (UTC)

To me, this seems like a clear case where having the conversions is excessive. I'd post the appropriate text of MOS:CONVERSIONS here, but someone's already done so. Also, I don't see why the lack of miles in, say, Canadian junction lists isn't being discussed as an "issue" here. If you're going to "suggest" junction lists using solely imperial units be saddled with metric clutter, I don't see why junction lists using just metric units shouldn't be saddled with imperial clutter. (As an aside, RJL-style, milepost-only tables have been brought through FAC 29 times, and if conversions for each milepost were truly mandatory, I'm sure someone would have required them by now.) – TMF 16:09, 2 August 2010 (UTC)

Both should be considered equally problematic. As a rule, all of these should either use miles converted into kilometres, or kilometres converted into miles, depending on which units is considered most appropriate to the country concerned (presumably miles first in the US, UK, Liberia and Burma, and kilometres first everywhere else). This allows people who use both systems to understand the articles equally. An American should not be disadvantaged when reading an article on a Canadian road, any more than a Canadian should be disadvantaged when reading an article on a US road.
I do not see any issue with converting an article such as Interstate 75 in Michigan or Pennsylvania Turnpike. Yes, the table has a lot of lines - but that shouldn't be a barrier. Note that {{convert}} does have a function for converting fields in tables.
That said, IMO we shouldn't have a problem with people getting creative when providing conversions in cases where space may be an issue. Part of the problem here is that those tables have no conversions whatsoever. Pfainuk talk 17:40, 2 August 2010 (UTC)
It's not so much the length, but the width of an already wide table that concerns me, Pfainuk. You'd want to shoehorn a 7th and 9th column into those tables when MOS:CONVERSIONS already exempts them from adding a conversion for every line in the tables. This issue was discarded in the past as noted above and at any of the 29 FACs where it was brought up. Category:FA-Class U.S. road transport articles lists the US highway FAs (Kansas Turnpike does not have an exit list table), and M62 motorway is another FA without conversions (although it should have them because of the UK-specific reasons detailed above). Imzadi 1979  18:07, 2 August 2010 (UTC)
I have a rather wide monitor, so it's perhaps unsurprising that I don't see a problem on that front. If there is a problem, then there are other ways of providing conversions. There's an article on speed limits somewhere that has a conversion table on the right hand side to save converting the same thing continually. You could also put a footnote on the column header saying something to the effect 1 mile = 1.6 km; 5 miles = 8 km, giving similar convenient versions for kilometres on articles on Canadian/Australian roads.
On M62 motorway, miles shouldn't just be included, they should almost certainly be the primary units as per practically all road signs in the UK. It seems perfectly illogical to insist on kilometres (and particularly kilometres only) based on a measure that most people probably couldn't even identify as distances in kilometres, when all the other signs give distances in miles.
Indeed, because of the failure to convert where a conversion is so clearly necessary - and the fact that the list of distances is not comprehensive - M62 motorway fails a reasonable reading of the FA criteria right now. Though I would note that the distances were added after the article made FA. Pfainuk talk 20:36, 2 August 2010 (UTC)
I could support adding a footnote with the conversion factors on the non-UK routes. --Rschen7754 20:55, 2 August 2010 (UTC)
Remember that the Driver location signs on all UK motorways are in kilometres. That is what drivers see and if they are plotting a route, they might well record that the exit where they turn off is at marker no x. In addition the Department for Transport is poised to publicise them - for example [2]. Martinvl (talk) 21:04, 2 August 2010 (UTC)
It is to driver location signs that I refer when I describe "a measure that most people probably couldn't even identify as distances in kilometres". You say the Highways Agency is poised to publish them based on a document that is three years old. If they were poised to do it when that document was published in 2007, it'd be all over by now. The fact that now, three years later, most people still probably couldn't identify them as distances in kilometres, would suggest that this effort to publicise them to the extent that you would like has failed. Miles remain the most appropriate unit in this context.
You say that people, when plotting routes, will record that the junction they're getting off is at driver location sign number x - a number that is only used on these signs and a few little-read documents - instead of the junction number, a number that is universally used in mapping and is far more clearly signed. This would seem distinctly bizarre and I would suggest that very few people are likely to go quite so far out of their way to make things complicated for themselves.
And none of this is a good reason not to use the mile as primary. The mile is by far the more commonly used unit in this context in the United Kingdom, and is thus clearly the more appropriate unit. Not including it at all is plainly absurd, and violates both WP:UNITS and, apparently, the local WikiProject's own guidelines. Pfainuk talk 21:44, 2 August 2010 (UTC)
(ec) Pfainuk, you misunderstand me, the UKRD project has developed a project-level guideline to use a column each for both miles and kilometers. The reason being that the driver location signs, and the logs of their locations, are based in kilometers while the rest of the signage and the units in use are in miles. It's my experience that UKRD is slow in updating articles to account for changes in standards. I too can live with such a footnote and leaving the need for an addition column out of the tables. Imzadi 1979  21:11, 2 August 2010 (UTC)
Fair enough. Though one would suggest that if their featured articles aren't up to scratch they probably should not be featured articles any more. As I note above, the driver location signs are an excuse for metrication, not a good reason. Pfainuk talk 21:44, 2 August 2010 (UTC)
Their lone FA was retained earlier this year at FARC. *shrugs* I wasn't aware of it before it was closed. Imzadi 1979  22:21, 2 August 2010 (UTC)
I have implemented a footnote in the templates that create the table headers for the Michigan articles, and added the necessary Footnotes section to the bottom of all the articles. Imzadi 1979  23:12, 2 August 2010 (UTC)

I'm not a fan of a footnote. For me, it makes the mile column unacceptably wide. It would also require the addition of a footnote section to every article that doesn't have one, and for the US alone, that number is in the several thousands. I also can't justify adding a section everywhere just for one short footnote. Perhaps a better option is to put the note in the table footer instead. – TMF 21:54, 2 August 2010 (UTC)

I too, don't like the footnote for the same reason as TMF. Would it be better to go the other way and place a sentence, such as "All distances are measured in miles. 1.000 mile = 1.609 km", above the list? For USRD articles, it could be added to {{Jcttop}} and its derivatives. Once that's done, all the articles using it are instantly updated. –Fredddie 00:51, 3 August 2010 (UTC)
Metric/imperial/US Customary conversions add to the clutter. However, many don't understand the old measures and metric measures are not universally understood, either. Therefore, both need to be provided in most contexts. Michael Glass (talk) 01:12, 3 August 2010 (UTC)
Then for all the people here pushing to change things, why hasn't anyone commented over this "issue" at Wikipedia:Featured article candidates/Don Valley Parkway/archive1 or any of the recent FACs for highway articles in the last year or more? Imzadi 1979  02:52, 3 August 2010 (UTC)

To all the people who are pushing for these useless metric conversions, I hope that you would be willing to update pages such as Interstate 5 in California, where the table is hardcoded, rather than just making more work for us editors who actually edit the road articles and who really don't see a need for metric conversions. If foreigners want to drive on American roads, they will have to deal with the US system of measurement already. --Rschen7754 04:20, 3 August 2010 (UTC)

I had converted the Michigan highway articles (all 217 of them) to specify a footnote in the template that creates the table header. Then I had to add a footnote section to the articles to get rid of the nasty red warning that there were references in the article that wouldn't display. After that was done, I did my grocery shopping this evening, made dinner and watched some TV. Now that I'm looking back over the articles, the extra note is making the column wider, and I'll be reverting the work tomorrow at some point. It's just messing up the table formatting. Honestly, MOS:CONVERSIONS does give an out here. MOS:RJL does not require the conversions. There have been 29 successfuls FACs in the past 3–4 years that have not had a second column or some other method. Sorry, I can't support this, and there is not the consensus here to make the change. Imzadi 1979  04:29, 3 August 2010 (UTC)

This is a perennial proposal; it was debated three years ago exactly. I agree with a lot of the points that were made there in opposition to requiring both mi and km. The perception that it results in "number soup" is one I share (I could see myself mixing up the columns upon reading them). I find Holderca's posts very relevant—as he says, the guide signs, odometers, and distance markers in the United States are all in miles, so the km conversions would be functionally useless to anyone attempting to apply them while visiting the road. km conversions are already listed in the infobox and route description, so users just wanting a conversion to "fix the distances in their head" would be adequately served by those. In any event, if this is forced through, I hope there's one of these pro-conversions people willing to apply the conversions to all 12,500 U.S. road articles. I doubt you're going to find anyone in the road project willing to spend their time doing something so menial resulting in such a marginal improvement to the articles. —Scott5114 [EXACT CHANGE ONLY] 05:11, 3 August 2010 (UTC)

  • As a big-picture issue, I've been gently pushing for road/highway/junction articles to be more accessible to international readers, and for editors from around the English-speaking world (even beyond) to foster collaborative arrangements. There are a number of significant advantages in that. I've left a few comments (by request) on Capitol Loop at Imzadi's talk page; these include suggestions that might increase readership by both non-experts, and by both US and non-US visitors. Avoiding metric conversions in tables is not helping to open the article up to a wider readership. Tony (talk) 05:26, 3 August 2010 (UTC)
That's not the issue though. How will taking the time and effort to add numbers that are of questionable utility increase readership? Why would people read an article that they'd otherwise be uninterested in just because we added km's? (Remember, if people aren't interested in a region or in transportation in general they're unlikely to read a road article unless it is of broad appeal like Route 66.) Why is the number of people reading a particular article of any concern, anyway? —Scott5114 [EXACT CHANGE ONLY] 05:44, 3 August 2010 (UTC)
I don't live in the US, Burma or Liberia; so I have to think hard to conceive "1.48 miles". It has an inward-looking feel about it, as well the (im)practical implications for non-US readers. And let's not encourage foreign road articles to remove US customary conversions. Tony (talk) 05:51, 3 August 2010 (UTC)
Keep in mind we're not talking about the whole article; we're talking about the junction list. The part where most readers with only a casual interest in the article would tune out. --Rschen7754 05:56, 3 August 2010 (UTC)
If you're just trying to "conceive" the distances, you're going to be looking at the prose (which has conversions) and the infobox (same).—Scott5114 [EXACT CHANGE ONLY] 06:45, 3 August 2010 (UTC)

Adding metric conversions to US articles will not increase their readership. The distances on the roads are based on miles, the speed limits are in miles, the government posts the lengths in miles (which if the fractions of lengths are converted to kilometres and added up, they probably would not be equal), and the cars show miles in larger print than kilometres. On the flip side, for every other country except basically the UK, metric is the ONLY system of measurement. I refuse to convert the Ontario articles to show imperial - Canada hasn't used in for 40 years, and 95% of the planet is in the same position. If you feel like converting all the distances at I-10 in Texas or Highway 401, be my guest... but I will probably revert it if it looks like crap and clutters the table with useless and repetitive information. - ʄɭoʏɗiaɲ τ ¢ 14:55, 3 August 2010 (UTC)

I'd prefer not to be called "stupid" in your edit-summary. So ... (1) do the tables have a different readership from the prose, or serve a different purpose? (2) should a table of convertible units in, say, an article on, say, the distances between US cities, or railways in New York State, also avoid metric conversions? Tony (talk) 15:08, 3 August 2010 (UTC)
I called the idea stupid. You should know this, you are the grammar whiz.
The conversion for the total distance is provided in the prose. The conversion for the distances to each exit are unnecessary and cluttering. WP:OTHERSTUFFEXISTS is great, but I don't care. - ʄɭoʏɗiaɲ τ ¢ 16:13, 3 August 2010 (UTC)
EDIT: And no, those converted units should be removed. The train map, the train conductor, the signs beside the train, the system map book you picked up at the station, and every other number you will see on that trip will ALL be in single unit of measurement. Converting it to km for lol's won't help anybody do anything at all. - ʄɭoʏɗiaɲ τ ¢ 16:20, 3 August 2010 (UTC)
On many US freeways, the distance in one or another set of units defines the exit number (Exit 136 is 136 miles from the start of the freeway or the state line). Having a second set of units in such circumstances would be confusing. In the case of junction distances lets just make it as easy as possible to implement WP:VERIFY and show one set of units - namely those that appear at the side of the road. Martinvl (talk) 15:25, 3 August 2010 (UTC)
Quite a few Canadian license plates can be spotted on New York roads. One can guess that a corresponding number of train passengers in New York are from Canada, so highway and railroad distances in New York should probably provide both American customary and metric distances. However, there is no need to provide conversions for speed limit signs, the Canadian drivers have no interest in those. Jc3s5h (talk) 15:21, 3 August 2010 (UTC)
No, there is no need to convert the distance between cities either if the American trains all use miles. Do you gain any benefit knowing the miles in another unit of measurement that isn't used anywhere else in that system? No! Even if you have used kilometres your entire life, you base distances and speeds on miles when you go to the States. - ʄɭoʏɗiaɲ τ ¢ 16:16, 3 August 2010 (UTC)

Let me make a good comparison using an example we all know. Angel Falls in Venezuela is the tallest single drop in the world. That drop is 807 metres of the 979 metre total drop. For me, this is fine and dandy. For an American, it means diddly, so we convert it to feet. AHHH! There we go, 2,648 ft out of a total of 3,212 ft. Makes sense now! By comparison, the distance I travelled along I-90 last spring through Pennsyvania was about 45 miles. The sign said it as soon as I got there. Did I sit there to convert it to km so I could say "Oh ok, now I can gauge that distance"? No, because the speed limit is in miles. If I converted one, I'd have to convert everything, but in the end I'd get the same result - About a 45 minute drive at the maximum speed limit. Knowing the km between two points measured in miles is pointless, because you can't use that converted value for any purpose. When you see the next sign saying "Ohio - 28", you're dealing with 28 miles, but you've already come 28 kilometres, but only 17 miles, so you have... hmmm... 28 miles to go until you're in Ohio. If the pure statistic of the kilometric distance is that important, you can pick an exit number and convert it. - ʄɭoʏɗiaɲ τ ¢ 16:31, 3 August 2010 (UTC)

The Economist style guide [3] says that conversions should be given the frist time in any article, but not subsequently. An example that they give is "It was hoped that after improvements to the engine the car would give 20km to the litre (47 miles per American gallon), compared with its present average of 15km per litre." You will notice only one conversion, not two (OK, I am not happy about expressing fuel consumption in km/L, but this is not the place to discuss that). My view is therefore is that if space is at a premium, then the first thing to sacrifice is the column of converted numbers. Martinvl (talk) 20:52, 3 August 2010 (UTC)
BTW, I see that M25 motorway provides conversions within the table. Works there ... Tony (talk) 15:14, 9 August 2010 (UTC)
Because both imperial and metric systems are used on the roads there. The UK tables are also designed fundamentally different from others (see WP:RJL), and have much more room with the lack of geographical context (location/division column). - ʄɭoʏɗiaɲ τ ¢ 15:25, 9 August 2010 (UTC)
Yeah, this is a red herring, the UK isn't following RJL at all. --Rschen7754 16:23, 9 August 2010 (UTC)

Timelines in the contents box

I was just wondering is there a MOS guideline on timeline dates in contents lists. 2010 Russian wildfires and also an older version of 2010 Northumbria Police manhunt highlight the problem. Depending on the date formatting used, the numbers look very messy when you end up with a list like..

2 Timeline

2.1 29 July

2.2 31 July

2.3 1 August

2.4 2 August

2.5 4 August

2.6 5 August

2.7 6 August

2.8 7 August

I would think this has come up before, but can not find anything stating how these sorts of things are meant to be handled. thanks BritishWatcher (talk) 21:16, 7 August 2010 (UTC)

How about
2.2 Saturday 31 July
2.3 Sunday 1 August
2.4 Monday 2 August
2.5 Wednesday 4 August
2.6 Thursday 5 August
2.7 Friday 6 August
2.8 Saturday 7 August
Martinvl (talk) 06:50, 8 August 2010 (UTC)
Yes that is possible, this is what was tried on the manhunt article..
2.2 July 31
2.3 August 1
2.4 August 2
2.5 August 3
Although it now just has a text title about the daily events. Showing the format in those ways is clearer, but is it ok just to use that sort of date format just for the contents, or would you then have to use it throughout the article to avoid two different formats? BritishWatcher (talk) 08:45, 8 August 2010 (UTC)
Except for Wikipedia articles that are explictly timeline articles, text headings are better, though I would have been tempted to use the following:
  • Birtley shootings (02:40, 4 July)
  • Denton shooting (14:00, 4 July)
  • Letter, sightings and appeals (5 July)
  • Cordon of Rothbury (6 July)
  • Further appeals, reward (7 July)
  • Stand-off with police and death (9/10 July)
BTW, looking through the article, I noticed that time were given using the 12 hour clock, but with "a.m." rather then the WP:MOS preferred "am". Also, I noticed one error where "a.m." was used when it should have been "p.m." (I corrected that). I would have use the 24 hour clock, especially as that is what the police use in their records. Martinvl (talk) 09:53, 8 August 2010 (UTC)
If "am" is "WP:MOS preferred", then my AWB edits are wrong. WP:MOS#Times says "dotted or undotted lower-case a.m. or p.m., or am or pm". Art LaPella (talk) 16:31, 8 August 2010 (UTC)
This is an inconsistency that has to be ironed out. I personally perfer "am" to "a.m.".Martinvl (talk) 19:45, 8 August 2010 (UTC)
It should be ironed-out. Some style guides I have seen indicate no space between time and small caps, no periods: 7:30AM and 7:30PM. 12-hour_clock#Typography indicates similar. It would be easy enough to create a template to do this. I prefer the typographical method without lower case and without periods. --Walter Görlitz (talk) 21:08, 8 August 2010 (UTC)
Indicating AM and PM is one thing that doesn't need to be uniformly standardized so long at it's clear. (There's still an infinity of articles with genuine ambiguities and perplexities that need to be clarified or corrected for the reader.) I don't like "am" for the simple reason that it's the first-person-singular form of "to be". [Modern British newspaper style also favours "eg" and "ie", but many Americans would have no idea what those mean.] 22:16, 8 August 2010 (UTC)
In my experience that's much of an ENGVAR issue, in that the capitals are more common in the US and the small letters in the UK. The periods and spaces are more common in the US than in the UK, though on both sides of the Atlantic they are getting rarer than they used to be. WP:RETAIN should apply. A. di M. (talk) 10:58, 9 August 2010 (UTC)

It's a sequence of events. The heading should emphasise the event name, the time of the event is a characteristic of the event and is secondary. Lightmouse (talk) 11:47, 8 August 2010 (UTC)

I agree with Lightmouse that timings are secondary and should therefore not be the main part of the heading. However that is not to say that there is no place for them in the headings - I believe their inclusion to be a matter of consensus of the editors concerned. Martinvl (talk) 12:34, 8 August 2010 (UTC)

Thanks. I also agree with you that the 24 hour clock is useful, particularly for data, transport, sequences, tables, timestamps. Lightmouse (talk) 12:38, 8 August 2010 (UTC)

Demonstration that YYYY-MM-DD format not ISO 8601

In a recent edit summary Wjemather stated "it is ISO format regardless of whether WP has adopted the standard or not. Added clarification". As a counterexanple, consider Microsoft Excel for Windows. It treats text entered into the current cell in the YYYY-MM-DD format as dates provided the date is on or after 1900-01-01; it treats earlier dates as text strings. Thus this widely used format is not equivalent to ISO 8601.

One facet of this topic where I have had no luck finding reliable sources is whether the people or government of any country ever adopted the YYYY-MM-DD format independently of ISO 8601, with different or unstated rules. I would appreciate being informed of any sources that address this.

P.S.: yes, I know about Excel's leap year difficulties. Jc3s5h (talk) 20:34, 12 August 2010 (UTC)

I think I need a clarification of the clarification. As I see it, User:Jc3s5h is claiming that just because a date is YYYY-MM-DD format it need not be because of adherence to ISO 8601 (it may have been "independently invented" i.e. without consulting or knowing about that standard) whereas User:Wjemather is claiming that if a date is in YYYY-MM-DD format it is ISO 8601 whether it is intentionally adhering to that standard or not. (And let's please leave aside what happens with leading zeros or whatever for the sake of simplicity.)
If that's the case it seems almost certain to me that it has been invented independently; it would be extremely surprising if nobody until those who framed ISO 8601 had ever thought of writing dates that way. Are we concerned particularly with the hyphens and number of digits or just generally with the endianness here? As for whether "the people or government of any country ever adopted the YYYY-MM-DD format independently of ISO 8601", that then very much depends on how specific you mean about that format (i.e. the hyphens, leading zeros, and so forth).
I don't see the relevance of how Excel treats user input. You can perfectly happily represent dates before 1900-01-01 in Excel (it's the OLE date format after all, with all its peculiarities), it just happens not to parse strings in a way that rejects dates outside of a certain range. I could test e.g. the Microsoft Windows API ::VariantChangeType to see whether that is an underlying feature of that API or specific to Excel, but I don't see how it matters to the argument anyway, since it manifests as a feature of the user input interface and has nothing to do with how dates are displayed or stored. Si Trew (talk) 20:58, 12 August 2010 (UTC)
I think the issue comes up when an article contains only post-1904 dates. In particular, the dates might appear in a table that was originally created in Excel and converted to Wikipedia format automatically. Then a new editor comes along, and wants to add some pre-1900 dates. But there is no guideline to tell the new editor what to do. If the editor is familiar with Excel and unfamiliar with ISO 8601, the editor might view the YYYY-MM-DD format as being peculiar to Excel, and since Excel can't easily handle pre-1900 dates the editor might feel the justification for that format no longer existed and change all the dates to dd Month YYYY format. If the editor is familiar with ISO 8601 and unfamiliar with Excel, he might use the YYYY-MM-DD format for the earlier format. Neither editor is wrong. An editor has just as much justification for supposing the YYYY-MM-DD format is the Excel format as the ISO 8601 format. Jc3s5h (talk) 21:14, 12 August 2010 (UTC)
If you look at any ISO standard, you will see that it was in use somewhere before it became an ISO standard - it only became an ISO standard because sufficient people were uising it to warrant the relevant ISO committees to sit aroud a table and agree a common standard. ISO 8601 evolved because it was being used in various countries like Sweden and South Africa (to mention but two), but more particularly because it was being used to timestamp electronic documents. I can vouch for it having been used in South Africa in 1972 - stamps in my passport prove that - and the ISO standard was first published in 1988. This does not make the South African useage non-compliant - in fact the South African and the Swedish useage was taken into account when drafting the standard.. Martinvl (talk) 21:30, 12 August 2010 (UTC)
Yes, I would agree that big-endian dates were obviously used for ages as a de facto standard before ISO 8601, and at least some of them are going to have used the YYYY-MM-DD format including the same spec. for hyphens, leading zeros, etc. The question was, can we get any RS that "people or governments" mandated it before ISO 8601 became a standard? Not that I think it really matters anyway: I'm not sure what good such an RS would do, really. Si Trew (talk) 22:34, 12 August 2010 (UTC)
As often seems to happen, MOS is hoist by its own petard here by defining a guideline for something without really defining what that something is, i.e. "YYYY-MM-DD format". Nevertheless, it does say that the only YYYY-MM-DD format acceptable is one compatible with ISO 8601. The earlier WP:DATESNO says:

YYYY-MM-DD style dates (1976-05-31) are uncommon in English prose [...] However, they may be useful [...] . Because it might be thought to be following the ISO 8601 standard, this format should only be used for dates expressed in the Gregorian calendar and for the years from 1583 through 9999.

(I've a slight annoyance with the year 1583 not 1752 here, but I doubt many YYYY-MM-DD dates are from before the last couple of centuries. They were encouraged by the {{cite/doc}} for that template's accessdate parameter, but no longer are; {{cite book/doc}} does show such a use in its examples, as does {{cite journal/doc}}, and I imagine others do too.)
So I see no trouble in either omitting or including the use of ISO 6801 in the sentence farther down in WP:MONTH (i.e. the one in the diff given by User:Jc3s5h at the top of this section). The disputed sentence:

Months are expressed as whole words (February, not 2), except in the ISO 8601 format (YYYY-MM-DD)

can keep or lose the "ISO 6801" whichever way the wind blows; if it's included, it's just a repeat of what's already said. The parentheses should be lost, though.
Si Trew (talk) 22:34, 12 August 2010 (UTC)
To largely echo that. MOS already states that YYYY-MM-DD formatted dates should comply with ISO 8601. Without going into all the issues, if it helps Jc3s5h's hypothetical Excel users, it is best that the term ISO 8601 is kept. Lose the parentheses by all means. wjematherbigissue 00:54, 13 August 2010 (UTC)
The only problem is using the term ISO 8601 format to relate to one date format referenced in the standard as ISO 8601 indicates. It would be like equating the spelling of color with the entire corpus of the "American English language". The standard defines more than short dates. So it would be acceptable to me to change the text to "Months are expressed as whole words (February, not 2), except in the YYYY-MM-DD as is seen in the ISO 8601 standard.".
It would also be incorrect to change the copy in the article while its resolution is being discussed. --Walter Görlitz (talk) 00:56, 13 August 2010 (UTC)
Some of the recent long discussions of date format make it evident that there are lots of editors around who are happy to use what they often call "ISO dates" but there is little agreement about the proper way to use the format for old dates, or whether variations such as 2010-10 (for August 2010) are allowed or not. Despite the plea in MOSNUM, there can be no assurance that the YYYY-MM-DD format is actually used in a way that conforms with ISO 8601.
In a more structured environment, I would propose a profile of ISO 8601, which would disallow representations such as 2010-10 and 20101012 (and disallow all the week number and period options). But in the Wikipedia environment I don't believe such a restriction could be enforced, so I think it is best to tell the truth: we don't know what editors intend when they write old or unusual dates that superficially follow the ISO 8601 spec, so we should put readers on notice that they have to figure it out from context. Jc3s5h (talk) 01:03, 13 August 2010 (UTC)
Your change made the guideline less clear, and almost invites people to use YYYY-MM-DD formatted dates that do not comply with the ISO standard. From what you have said, I don't see why you would want to do that. What exactly are you suggesting now? wjematherbigissue 01:14, 13 August 2010 (UTC)
First, how can someone create a YYYY-MM-DD formatted date that does not comply with the ISO 8601 standard? The reverse is possible: to use an ISO 8601 date format that is not YYYY-MM-DD (YYYYMMDD for instance). Second, to which change are you referring? The one in the article or the comment made above yours? --Walter Görlitz (talk) 01:32, 13 August 2010 (UTC)
Any date before Oct 15, 1582 written in the YYYY-MM-DD fails the ISO format. --MASEM (t) 01:37, 13 August 2010 (UTC)
Not as such, the ISO 8601 article states (I've not looked at the spec itself) "Earlier dates, in the proleptic Gregorian calendar, may be used by mutual agreement of the partners exchanging information. The standard states that every date must be consecutive, so usage of the Julian calendar would be contrary to the standard (because at the switchover date, the dates would not be consecutive)". So dates written for before then can be compliant with ISO 8601, but would not conform with WP:DATESNO anyway.
Anyway, I can think of a YYYY-MM-DD format that doesn't conform: The year is written as a four-digit decimal number, padded with leading zeros if necessary; the day is written as the number of days until the start of the next month, like the Romans did, padded with a leading zero if necessary, and the month is written as a two-letter abbreviation JA, FE, MR, AP, MY, JN, JY, AU, SE, OC, NO, DE. The three are written together and separated by unspaced hyphens.
Of course I could make up very many other such formats. That's what I mean by MOS not really defining what it means by YYYY-MM-DD format. Si Trew (talk) 06:43, 13 August 2010 (UTC)
This was the original change by Jc3s5h I am referring to. wjematherbigissue 01:43, 13 August 2010 (UTC)


Just a side comment that I feel I should make. It is only my opinion, but this whole argument is getting kinda lame in my book. Call the damn things "ISO style date format" and move on. The usage may not be 100% compliant with the applicable ISO standard nor have we formally adopted the standard. Personally, I think the things should be excised from display text except where appropriate. Use them to set up sorting in tables using {{sort}} and display the content in another format, but get rid of them in most other cases. Running prose and footnotes should be using a single, consistent date format that's used in the sources appropriate to the topic. Rarely is space a concern in footnote formatting, unlike some tables. I can see using ISO style dates in some topics, but they really should not be in use in many topics. Imzadi 1979  01:17, 13 August 2010 (UTC)

To answer Walter Görlitz's question, if I were to rewrite the lead sentence of Guy Fawkes Night to read

Guy Fawkes Night, also known as Bonfire Night, is an annual celebration held on the evening of 5 November to mark the failure of the Gunpowder Plot of 1605-11-05.

that would contain a date in the YYYY-MM-DD format that does not comply with ISO 8601. Jc3s5h (talk) 04:11, 13 August 2010 (UTC)
And so not conform to WP:DATESNO which states that dates thus written should be in the Gregorian calendar. That's my quibble with WP:DATESNO saying don't use it for dates before 1583 (the year it was introduced in most Roman Catholic countries) instead of 1752 (the date it was introduced in Great Britain and her Colonies) since it seems to let slip dates between those years, although it says it should use Gregorian dates and so presumably one would have to write 1605-11-18 (is that right?) which is absurd. Of course, I assume it's taken as common sense that articles discussing the date formats themselves are exempt without having to call them out.
The reason I suggested losing the parentheses – "ISO 8601 format YYYY-MM-DD" – was so as not to imply that ISO 8601 defines only that one format. (If "ISO 8601 goes, the parens have to be lost anyway.) Si Trew (talk) 06:23, 13 August 2010 (UTC)
As I said earlier, it is not a good idea to lose ISO 8601 from the wording since it could lead to the ambiguity demonstrated above. We know readers often go straight to the section of MOS they are looking for, so may easily miss the earlier statement that YYYY-MM-DD should be ISO compliant. I also think it is helpful to reiterate the accepted ISO format. As such, I support your suggestion of "ISO 8601 format YYYY-MM-DD", no parentheses. wjematherbigissue 09:31, 13 August 2010 (UTC)
I've changed {{cite journal/doc}} and {{cite book/doc}} not to use the YYYY-MM-DD format, and checked {{cite news}}, {{cite/doc}} ({{Citation/doc}}), {{cite web/doc}} and {{cite thesis/doc}} which were OK anyway. Any others? Si Trew (talk) 06:54, 13 August 2010 (UTC)
... I have revoked these changes as the discussion has not yet been completed. Martinvl (talk) 11:43, 13 August 2010 (UTC)
I think you should reconsider that. Some of the {{cite}} templates used to stipulate YYYY-MM-DD format, now they do not (the ones I mentioned state that they should use the format used in the rest of the article.) I can't find where in the history this was changed, but I assume that the stipulation for YYYY-MM-DD was removed. In that case, considering that MOSNUM already discourages using this format, and that it is only the examples that uses it and not the description of the parameter, I imagine it was just an oversight not to change the example. In any case, I don't see how it pre-empts the outcome of this dicussion here; I mentioned the cite templates as an aside; this discussion is about whether the term "ISO 6801" is required or not. Si Trew (talk) 15:33, 13 August 2010 (UTC)

The current state "the ISO 8601 compliant YYYY-MM-DD format" is perfect. --Walter Görlitz (talk) 13:58, 13 August 2010 (UTC)

I don't understand. The phrase stated by Walter Görlitz does not appear in WP:MOSNUM. Jc3s5h (talk) 14:26, 13 August 2010 (UTC)
It was added and then reverted with comments about potential lynch mobs. I think it addresses both concerns. --Walter Görlitz (talk) 14:30, 13 August 2010 (UTC)
"The ISO 8601 compliant YYYY-MM-DD format" may be OK if the reader actually takes the trouble to read the ISO 8601 article. But a hasty reader who does not read the article might think that all he/she has to do to comply with ISO 8601 is use the YYYY-MM-DD format. Jc3s5h (talk) 16:17, 13 August 2010 (UTC)
But is the guideline that we must meet ISO 8601 or that we should format dates as YYYY-MM-DD? As has been shown, ISO 8601 standard does not work under all situations while YYYY-MM-DD does. --Walter Görlitz (talk) 19:16, 13 August 2010 (UTC)
Is it not possible to add a phrase to the effect that "Strictly speaking, only the Gregorian Calendar can be ISO 8601 compliant". This should satisfy everybody.Martinvl (talk) 20:06, 13 August 2010 (UTC)
Suppose we held a well-advertised RfC on whether we should require that whenever all-numeric dates are used in Wikipedia articles, they must comply with ISO 8601, and must be in the form YYYY-MM-DD. I expect that the responses would be all over the map. Some would want to ban all-numeric dates completely, some would say yes, do exactly what is proposed, some would say that all-numeric dates must be interpreted in context and it is a mistake to have the 1583 and Gregorian requirements in MOSNUM. I doubt any consensus would emerge, and I think the present version of the MOSNUM exists because not too many people have noticed what it says about all-numeric dates. Jc3s5h (talk) 20:26, 13 August 2010 (UTC)

Please permit me to take this conversation a step backward. The statement is about not padding month numbers except when used in YYYY-MM-DD format. How can we word it so that we exclude this ISO 8601 standard debate? --Walter Görlitz (talk) 21:35, 13 August 2010 (UTC)

Terminology for quantity ranges

IMO "from N1 to N2" and "N1..N2" should be preferred over "between N1 and N2" to avoid ambiguity, since "between N1 and N2" can be used colloquially to men either "greater than N1 and less than N2" or "equal or greater to N1 and less than or equal to N2" is used colloquially to mean could mean "between N1 and N2 exclusively" or "between N1 and N2 inclusively". It's generally not a problem with narrow ranges (for example it would be unusual for someone to use "between 2 and 4" as a verbose way of saying "3"), however it's less clear for broder ranges (for example the corvidae article contains the statement that "Corvids can lay between 3 and 10 eggs", which could be interpreted as either a needlessly awkward and ambiguous way of saying "3 to 10 eggs" or an even more needlessly awkward and ambiguous way of saying "4 to 9 eggs"). -- Gordon Ecker (talk) 06:37, 11 August 2010 (UTC)

I understand your point, which I find particularly tricky with "Since [year]". But one must also realize the limits of the Manual of Style and its dozens of ramifications (such as MOSNUM). How many editors read significant parts of them even now? (They usually see only a single point that's mentioned in an edit notice or talk page.) How many more points can we stick into an MoS and realistically expect editors to notice (or even remember once they've read them)? This isn't to attack your point specifically, but to wonder how many ambiguities and obscurities can be cured by mentioning them in the Manuals of Style, and how many have to be tackled on the fly as articles are read and edited. ¶ On the other hand, a genuine ambiguity or obscurity (if indeed "from ... to" is really unambiguous) is more deserving of treatment in an MoS than some stylistic point preferred purely for reasons of aesthetics or uniformity, e.g., whether or not to put a full-stop/period after m, in, ha, or gal — or when to use the em dash or serial comma, as I just did for clarity, or indeed how to punctuate e.g. —— Shakescene (talk) 01:11, 17 August 2010 (UTC)

Non-breaking space

Use a non-breaking space (also known as a hard space) to prevent the end-of-line displacement of elements that would be awkward at the beginning of a new line:

  • in other places where breaking across lines might be disruptive to the reader, such as ... 7 World Trade Center

When would you ever write "7 World Trade Center"? For one, "7" should be spelt out and two, why would you need a non-breaking space there? McLerristarr (Mclay1) (talk) 12:44, 13 August 2010 (UTC)

As an address, it would not be spelled-out, at least not on America (or Canada). --Walter Görlitz (talk) 13:56, 13 August 2010 (UTC)
A British example would be "10 Downing Street". --Walter Görlitz (talk) 15:44, 13 August 2010 (UTC)
That being said, "7 World
Trade Center" or "10 Downing
Street" are probably worse line breaks than breaking after the number. Si Trew (talk) 17:17, 13 August 2010 (UTC)
The guideline never stipulates that only numbers and an associated word be connected by a non-breaking space. I often connect several words by non-breaking spaces in an instance where a word would seem odd without its counterpart if separated by a line break. For example, the Kawartha lakes as opposed to the Kawartha
lakes
- ʄɭoʏɗiaɲ τ ¢ 17:48, 13 August 2010 (UTC)
Sure, but if the browser can't actually display the whole text on one line it will break it at the non-breaking space anyway (I assume); in this case it might actually be a disadvantage to put the non-breaking space in between the number and the street name since it would make it less likely to break it there, and actually to break it at your "non-breaking space" between the words. I'm no expert in the requirements from HTML of how browsers decide this, and would imagine they are allowed a lot of freedom to choose which of consecutive "non-breaking spaces" can, in fact, be treated as a breaking space. ('d hope so, since breaking at the first space is not always the best policy, see e.g. Knuth, Donald E. (March 1980). "Mathematical Typography". Dr. Dobb's Journal of Computer Calisthenics and Orthodontia. 5 (3). People's Computer Company: 9. ISBN 0-8104-5492-0. Si Trew (talk) 14:57, 14 August 2010 (UTC)
I had no idea it was an address. I think we should change the example, especially since that address doesn't exist anymore, does it? McLerristarr (Mclay1) (talk) 15:07, 14 August 2010 (UTC)
Is there a need for only current addresses to be listed? It could be used in historic articles. --Walter Görlitz (talk) 16:00, 14 August 2010 (UTC)
Perhaps switching it to 10 Downing Street instead. I think that's an equally famous address, but it's much more apparent that it is a street address and not a building name. - ʄɭoʏɗiaɲ τ ¢ 16:32, 14 August 2010 (UTC)
I use that rule in my AWB edits, which probably makes me the primary person you need to communicate with, as most editors never read this far. I interpreted "7 World Trade Center" to mean one-digit addresses only, and so I didn't try to program for it because that hardly ever happens. Today WP:NBSP says "123 Fake Street" which probably happens more often, mostly in citations, assuming I abstract it to "digits – any word or 2 words – street or avenue". So does the number of digits matter? Art LaPella (talk) 17:54, 14 August 2010 (UTC)
I'd say no, it doesn't matter. 10
Downing Street is likely to be as distracting as 1600
Pennsylvania Avenue or 7
World Trade Centre. Pfainuk talk 19:40, 14 August 2010 (UTC)

NBSP and hyphens in citations

There is a discussion at User talk:Art LaPella#Your AWB edits concerning whether WP:NBSP should be applied within date parameters of a citation template as in date={{Nowrap|6 November}} 2010. It also concerns whether hyphens within titles should be changed to dashes according to the WP:DASH rules that apply elsewhere, as in this previous discussion. LeadSongDog come howl! 15:10, 19 August 2010 (UTC)

Surely it would be more productive and less clutterous to handle this from inside the template, something like {{#time:j" "F Y|{{{date|}}}}} in this. Try it. See also mw:Help:Extension:ParserFunctions##time. ―cobaltcigs 16:45, 19 August 2010 (UTC)

I don't understand it enough to try it, but would it change mm dd, yyyy to dd mm yyyy? If so, wouldn't that cause an WP:ENGVAR issue? Art LaPella (talk) 19:24, 19 August 2010 (UTC)

With appropriate conditionals you could make it follow the order of the input. A rudimentary attempt might look like this:

{{#switch:{{{date|}}}
| {{#time:j F Y|{{{date}}}}}  = {{#time:j" "F Y|{{{date|}}}}}
| {{#time:F j, Y|{{{date}}}}} = {{#time:F" "j, Y|{{{date|}}}}}
| #default = {{{date|}}}
}}

Beyond that you could add a few more cases to catch abbreviated months, leading zeroes, missing or extra commas, and other common deviancies to achieve greater immediate penetration (fewer defaulting cases requiring preparatory clean-up).

Alternatively you could add a {{{df}}} parameter like those already present in {{birth date and age}} and similar templates. This would simplify the code somewhat while allowing the {{#time}} func to catch nearly anything and reformat it according to {{{df}}}:

{{#if:{{{df|}}} <!-- day first -->
| {{#time:j"&nbsp;"F Y|{{{date|}}}}}   <!-- 6&nbsp;November 2010 -->
| {{#time:F"&nbsp;"j, Y|{{{date|}}}}} <!-- November&nbsp;6, 2010 -->
}}

Yes, the ideal solution would be a way for one template to know which other templates are being used on the same page (namely whether they include {{mdy}}/{{dmy}}) and make decisions based on that, but it’s not possible with current software. Look on bugzilla for something about that. The next best solution may be for bots to add the {{{df}}} parameter to cite_foo template calls based on which format is declared on the page, but hey—if you already had planned to edit a bunch of pages… ―cobaltcigs 20:53, 19 August 2010 (UTC)

The nbsp in November 6 might be controversial, but easily removed. Art LaPella (talk) 21:56, 19 August 2010 (UTC)
I’m not convinced any of them are necessary or helpful, but I do see whatever puts them furthest out of sight and mind (while achieving more consistency with fewer edits) as a good thing. ―cobaltcigs 22:32, 19 August 2010 (UTC)
I have expressed a similar opinion. Art LaPella (talk) 23:02, 19 August 2010 (UTC)
Agreed. An examination of the text at Template:Citation/doc#Dates and the code in Template:Citation is instructive. While it has a |dateformat= it is unclear (at least to me) what function it performs towards defining the |Date= parameter passed to Template:Citation/core. LeadSongDog come howl! 16:13, 20 August 2010 (UTC)
If this causes endless programming problems, I wonder if the style people would object to something simpler: {{Nowrap}} the whole date field and be done with it? Art LaPella (talk) 17:33, 20 August 2010 (UTC)

I don’t believe the problems would be “endless”, but no-wrapping the entire parameter automatically within the template would be saner than no-wrapping the input manually in each article. I’d consider both options less than ideal, but the internal approach requires fewer edits and will be easier to adapt to whatever new software features come available. ―cobaltcigs 01:06, 22 August 2010 (UTC)

OK, my experience in changing templates is limited to changing punctuation in simple text fields. For the detailed option described above, would it shake your confidence to realize that citation template date fields might have anything from my Nowraps (which I have stopped adding), to misformatted dates I've seen like 07=14-06 (note equals sign), to any other unformatted nonsense that might be lurking there? Or would all such citation templates instantly transform into that big ugly CITE ERROR message? Art LaPella (talk) 01:26, 22 August 2010 (UTC)

If you add the #iferror: function to the second example, you can default to the status quo when facing unformatted nonsense:

{{#iferror: <!-- return the following unless it produces an error… -->
  {{#if:{{{df|}}} <!-- day first -->
  | {{#time:j"&nbsp;"F Y|{{{date|}}}}}   <!-- 6&nbsp;November 2010 -->
  | {{#time:F"&nbsp;"j, Y|{{{date|}}}}} <!-- November&nbsp;6, 2010 -->
  }}
| {{{date|}}} <!-- …in which case, act like nothing happened -->
}}

#iferror is not necessary in the first example as an error merely would cause the switch statement to #default. The disadvantage to the first is that it re-formats based on an exact match. The second example re-formats anything resembling a date, but requires instructions as to which format. A third way would benefit from some function to tell us quite simply the order in which the month and day appear in the input, or one which tells us which date format should be used in the article at large (by asking whether it transcludes {{dmy}}/{{mdy}}, etc.). ―cobaltcigs 01:52, 22 August 2010 (UTC)

Sounds good except for the word "you". If my first such experiment were on a template invoked 6.37 bazillion times, with the added complication of how the core template relates to the, um, un-core templates, then I assure you the templates would be in more peril than anything I can mess up using AWB! Do we have any volunteers? Art LaPella (talk) 02:49, 22 August 2010 (UTC)

If it isn't reverted, this edit probably, though not unambiguously, means that all date fields should be entirely Nowrapped. The edit only mentions a dd month yyyy date, leaving us guessing about month dd, yyyy. I think there is a parameter that does the same thing as Nowrap. The same should be done to the {{date}} template itself. Art LaPella (talk) 22:35, 22 August 2010 (UTC)

  • These templates are bad news. They seriously detract from the "anyone can edit" pillar. They are complex. They are much better input via plain text. Tony (talk) 23:07, 22 August 2010 (UTC)
    • I don't think there's a consensus for that, but let's pretend there is. After the templates are burned, should it be 6&nbsp;November&nbsp;2010 and November&nbsp;6,&nbsp;2010, even in a citation? Art LaPella (talk) 04:12, 23 August 2010 (UTC)
      • Perhaps Tony might have a point if we had an "all editors must use templates to cite" rule, but we don't so he doesn't. We collaborate. Many editors leave only vague hints where they found their sources, yet other citegnoming editors manage to track down the source from those hints and provide full citations, in whatever format the article already uses. Rather than worry about whether every editor has memorized the entire MOS, how about we just get on with making it easier to do the right thing? LeadSongDog come howl! 20:22, 23 August 2010 (UTC)

The price of plain text is inconsistent order of information and ambiguity with respect to which text is supposed to represent what. I think ideally we’d have a database table to catalogue reference citations, and thus be able for auditing purposes to pull up an immediate and exhaustive list of articles whose content depends upon a specific source, be it the Los Angeles Times or Poor Richard’s Almanack, with dates, authors, etc. in a similarly machine-readable format. For now all we have is the parameter list where each template is invoked. Some readers may enjoy having non-breaking spaces, spans of style="white-space:nowrap;" and other formatting quirks in the html output, but including them as part of the input (wiki-text) only will complicate attempts to parse said input. Same principle applies to infoboxes, succession chains, etc. as far as I’m concerned. ―cobaltcigs 06:12, 23 August 2010 (UTC)

Also, to argue in an nbsp thread that citation templates are incomprehensible is a fine example of the Manual of Style subculture gap I'm trying to bridge (even if Nowrap in a date field was the wrong way to do it). The rest of Wikipedia considers nbsp to be more incomprehensible than citation templates. Citation templates are everywhere, but when I find an nbsp, it's likely to be my own. Art LaPella (talk) 14:11, 23 August 2010 (UTC)
(And having a simpler way to input non-breaking spaces is something which will never be done, probably because someone wants the gap to stay. A. di M. (talk) 15:48, 23 August 2010 (UTC))
There was a suggestion that the software should interpret the underscore character as a non-breaking space, at the Village Pump a month or two back. That struck me as a very good idea. Not sure how the discussion ended up; as often the way with such suggestions there, my impression was that it got lost between various technical and non-technical schools of thought and didn't get anywhere, but I could be wrong. PL290 (talk) 16:20, 23 August 2010 (UTC)
I'm impressed by the number of sensible suggestions that cobaltcigs has come up with. To {{nowrap}} the |date= fields in citation templates sounds like a sensible step. Art LaPella's recent edits to add lots of {{nowrap}}s are in the right intention, but it's a 3,000,000 edit solution versus a single citation/core edit solution, and more nested templates in citations doesn't make them easier to read (maybe if citation/core is updated we can remove those manually added nowraps again). It's true that there are various incorrect formats in use out there, part of my contributions to WP:AWB is to catch & tidy them. If there are more fixes to citation dates that AWB could make, let me know. Rjwilmsi 23:01, 28 August 2010 (UTC)
What's needed here is 1) specifications from the style people. Do we Nowrap the dates no matter what, or only if they are day month year? The former would be easier. And 2) implementation from the template people. Art LaPella (talk) 23:42, 28 August 2010 (UTC)
Add to that 3) Do we nowrap dates at all in citations, or only in body text? LeadSongDog come howl! 03:48, 29 August 2010 (UTC)
If you're asking me, no, I won't make such edits again. If you're asking the style people about changing the template, probably yes, but it's true they haven't answered that either. Art LaPella (talk) 04:28, 29 August 2010 (UTC)
Thanks for pointing that out. I should have been clearer. Yes, I'm asking the style people, after all, this is that talkpage. I kind of AGF'd that you wouldn't be making those edits in the interim, but thank you for making it explicit. Let's hope we can get some clarification on the Q's fairly soon. LeadSongDog come howl! 04:38, 29 August 2010 (UTC)

Merge requests

There are several similar unit articles. Please see the merge discussions about the following unit articles:

Regards Lightmouse (talk) 12:26, 20 August 2010 (UTC)

Numerals for centuries

Rich, is there some reason the guidance should be changed to allow the spelling out of the number? Tony (talk) 00:47, 22 August 2010 (UTC)


Yes, there are a number of reasons that we should revert this change - which was a revert of this, which is a revert of this which changes the initial "may be expressed as numbers" introduced by you as part of your sweeping re-write in 2007, and later changed to "are". Prior to this the advice was explicit

  • 18th century or eighteenth century.

As introduced by me, here in February 2006, and standing for a long time (apart form a brief words-only interregnum) pursuant to this discussion.

The previous apparent implicit advice to use numbers actually started out under this rubric:

A page title that is just a number is always a year. Pages also exist for days of the year, decades, centuries and even millennia. The formats are:

In other words a description of page titles became, in time a prescription for a text style.

Having laid out the wiki-history (there are also historical discussions - see the archive) let us look at style:

Fist, usage - in books the words are almost always spelled out - picking up the two on my desk "Who's Who in the Bible" and "Flowers in Britain" both use this style - and the latter has many numerals in its text in the form of measurements.

Secondly style guides either support words or are agnostic.

  • Chicago Manual of Style 9.36 (Online) supports or (Chicago 8.40) "should be spelled out in lowercase letters (e.g. ninth century, twentieth century). Spell out decades (the sixties, the seventies) or if the decade is identified by the century, write them as plural numerals (1920s, 1880s)."
  • MLA 3.5.5 (6th ed.)) supports
  • Oxford manual of style, "Oxford style is to use words. (p. 191)"
  • Possibly AP may differ, but we are not a newspaper.

Third, barbarism.

  • Avoid starting sentences with numerals - therefore we want to be able to say "Thirteenth-century pottery is in particular demand in Greenwich." The same applies to headings.
  • Hyphenating numerals and words is akin to using a metric nut on an imperial bolt: constructs such as "12th-century" should be eschewed at all costs.

Fourth, consistency.

  • We use spelled out ordinals for almost everything, apart form direct quotes there are very few exceptions, and we use spelled out cardinals for small numbers too (I think Chicago's two word rule is good here).
  • Personally I would like to see words only for centuries, but the complaints about "to many keystrokes" would doubtless be deafening.

Fifth meaning - contrary to obvious expectation a century is not a mathematical construct, certainly there is a technical meaning 00:00 on the first day of 1901 to 24:00 on the last day of 2000, was the twentieth century, however it is also use to describe an era and can significantly over-lap or under-lap one or both of the official ends of the century, using a numerical format does not convey this meaning as well.

Sixthly layout - numerals disrupt the flow of the text, more so in some typefaces than others. And while they are welcome in scientific and technical contexts, which tend to be replete with tables and graphs, in discussions about humanities and arts they should be firmly kept in check (speaking as a mathematician).

Regards, Rich Farmbrough, 05:07, 3 September 2010 (UTC).

I disagree with just about everything in your post. Do you have modern editions of style guides, and of a wider range of them? Tony (talk) 05:35, 3 September 2010 (UTC)

Here's few:

Regards, Rich Farmbrough, 16:42, 4 September 2010 (UTC).

Planets and such

I've noticed most, if not all, of the articles on the planets and moons of the solar system do not give conversions in Imperial/US customary units. Is there a particular reason for this, and where is it covered in the MOS that this is permisible? Given that the bulk of EN.WP's readership is still in the US, this seems quite odd. Thanks. - BilCat (talk) 07:49, 31 August 2010 (UTC)

American scientists use the metric system. I suppose that's the reason. McLerristarr / Mclay1 07:56, 31 August 2010 (UTC)
See MOS:CONVERSIONS, the bullet starting with "When units are part of the subject of a topic". But I think that at least the lead section should have conversions (or else the first occurrence of each unit should be linked, or there should be a footnote with a conversion factor, or whatever works), per WP:MTAA. A. di M. (talk) 09:17, 31 August 2010 (UTC)
OK, thanks for the links. - BilCat (talk) 11:21, 31 August 2010 (UTC)

Currencies

"When there are different currencies using the same symbol, use the full abbreviation ... unless the currency which is meant is clear from the context." Using the same symbol in the world, or using the same symbol in the article? If we link and define "$" as the US dollar in an article that uses US dollars, is "$" good enough to represent the US dollar, even if that symbol means something else in a country closely tied to the article? (In this case, Argentinians use "$" for their peso and "US$" or similar for the U.S. dollar.) - Dank (push to talk) 13:15, 20 August 2010 (UTC)

Yes, this is the English, not the Spanish WP. The US dollar is the dominant currency in the world, let alone among English speakers. It is clutter, I think, to insert "US", unless it's an article in which other "dollar" currencies—such as those of Australia, NZ or Canada—are at issue, where the distinction needs to be made during the article whenever unclear. I believe there is utterly no reason to link to either "US$" or "$". Argentina-related articles should not, I think, use a dollar sign to represent their peso. That would be fine in Spanish, and if in a direct quote here, should be disambiguated in square brackets. Tony (talk) 00:51, 22 August 2010 (UTC)
Thanks much, agreed. I inserted "in the article", which I think was meant already, just to make sure people don't think we mean "generally in Wikipedia" or "generally in the world". - Dank (push to talk) 02:28, 22 August 2010 (UTC)
I'm not sure I'd always sacrifice clarity to save two bytes. Most people wouldn't know whether or not the (e.g.) Zimbabwean currency is called dollar, so in an article about Zimbabwe they might not know whether $ must be the US one or could be another one. (OTOH, once you say US$ once you can just use $ thereafter unless the article deals with several different $s). —Preceding unsigned comment added by A. di M. (talkcontribs) 23:08, 22 August 2010 (UTC)
I think it's unfair to say the American dollar is the dominant currency in the world. There may be more Americans than other native English-speakers but the pound and euro are worth more and are just as important in the world market. Australians, Canadians and New Zealanders also use dollars, so it's unfair to assume that everyone will always assume the article means American dollars. If the context has nothing to do with America, then US$ should be specified. McLerristarr / Mclay1 08:01, 31 August 2010 (UTC)

I have amended the point and put "in the article" back in. McLerristarr / Mclay1 04:47, 12 September 2010 (UTC)

Thanks. - Dank (push to talk) 15:11, 12 September 2010 (UTC)

Percentage harmony with MoS

MOSNUM and MoS are pretty similar, but there's a minor niggle, and it would be better to make the wording and order the same where there's an opportunity:

MOSNUM

  • Percent is commonly used to indicate percentages in the body of an article. The symbol % is more common in scientific or technical articles and in complex listings.
  • The symbol is unspaced (71%, not 71 %).
  • In tables and infoboxes, the symbol % is normally preferred to the spelled-out percent.
  • Ranges should be formatted with one rather than two percentage signifiers (22–28%, not 22%–28%).
  • Avoid ambiguity in expressing a change of rates. This can be done by using percentage points, not percentages, to express a change in a percentage or the difference between two percentages; for example, The agent raised the commission by five percentage points, from 10 to 15% (if the 10% commission had instead been raised by 5%, the new rate would have been 10.5%). It is often possible to recast the sentence to avoid the ambiguity (raised the commission from 10% to 15%.). Percentage point should not be confused with basis point, which is a hundredth of a percentage point.

MoS

  • Generally, use either percent or per cent to indicate percentages in the body of an article.
  • The percentage symbol (%) is preferred in scientific or technical articles, in complex listings, and in articles where many percentages are reported.
  • Do not put a space before the symbol (71%, not 71 %).
  • In tables and infoboxes, use the percentage symbol, not the words (71%, not percent or per cent).
  • For ranges, use one percentage symbol, not two (22–28%, not 22%–28%).

Comments

  • Why it is necessary or desirable in MOSNUM to imply that non-scientific articles are less likely to use the symbol?
  • Ranges looks simpler and better for MoS.
  • Unspaced looks better in MOSNUM.
  • A brief mention (with a see MOSNUM) might be made of the percentage-point issue at MoS.
  • Do ANY infoboxes or tables spell out "per cent" or "percent"? I think there's a case for insisting on the symbol. Both locations, by their nature, need to save on space—it's part of the whole point. Tony (talk) 01:56, 10 September 2010 (UTC)
To answer the percent question, I didn't find "percent" or "per cent" in infoboxes or in the body of a table. I estimate "percent" in table headings to occur in about 2500 articles. Examples are Chemical makeup of the human body and Abundance of the chemical elements. The word "Percentage" is more often used the same way. In a heading labeled "Percent", changing it to just "%" would look strange. Art LaPella (talk) 05:19, 10 September 2010 (UTC)
Personally I'd like to see the main MoS spell out the problems with expressing rate changes. It's just such a common cause of avoidable confusion and hysteria (Hormone replacement therapy increases the risk of invasive breast cancer by 26%!!!). Adrian J. Hunter(talkcontribs) 09:50, 10 September 2010 (UTC)
Indeed, I had a jolly time insisting on this distinction in the Australian elections pages. Margins, swings, are points, not percentages. Tony (talk) 11:31, 10 September 2010 (UTC)

Longer Periods - Centuries

The text used to read: Forms such as the 1700s are normally best avoided (although the difference in meaning should be noted: the 1700s is 1700–1799, whereas the 18th century is 1701–1800). 1700s is ambiguous, as it could also be confused for the first decade of the 1700s, 1700–1709.
This was changed to: Forms such as the 1700s are normally best avoided (although the difference in meaning should be noted: the 1700s may mean 1700–1799, whereas the 18th century is 1701–1800) 1700s is ambiguous, as it could also be confused for the first decade of the 1700s, 1700–1709. (The change being from "the 1700s is 1700–1799" to "1700s may mean 1700–1799")

Given the context is perfectly clear (highlighting that the 1700s and the 18th century are not the same thing, i.e do not cover the same years), 1700s can only mean 1700–1799, so the change to "may mean" is unnecessary at best and possibly just plain wrong. My reversion to the previous wording has been contested, so I seek clarification. Perhaps it would be best to reword it altogether. wjematherbigissue 09:30, 3 September 2010 (UTC)

When making the change you reverted, User:Hmains pointed out the reason: it was to align that statement with the later statement, "1700s is ambiguous, as it could also be confused for the first decade of the 1700s, 1700–1709." Your revert reintroduced that inconsistency. The question of what to call the first (and second?) decades of a century seems to be perpetually debated, but should we conclude that the MoS should rule "may mean" to be "plain wrong", corresponding attention would be required to the end of the passage. PL290 (talk) 09:46, 3 September 2010 (UTC)
No "alignment" is necessary and there is no inconsistency. In the context it is used, the 1700s is 1700–1799. It can mean nothing else. If that is somehow unclear, then maybe we should focus on coming up with better wording. How about this:
Forms such as the 1700s are normally best avoided since it may be unclear whether a 10 or 100 year period is meant (i.e. 1700–1709 or 1700–1799). It should also be noted that the 18th century (1701–1800) and the 1700s (1700–1799) do not span the same period.
Thoughts? wjematherbigissue 10:30, 3 September 2010 (UTC)
Taking the original sentence in isolation, I would agree with your view of it; however, it is not an isolated statement. Its context is its paragraph. Your suggested new wording is, in my opinion, an adequate solution, and a definite improvement on the current wording. PL290 (talk) 12:30, 3 September 2010 (UTC)
Agree with PL290. "1700s is 1700–1799 ...1700s is ambiguous" is a contradiction, and the rewrite corrects it. Art LaPella (talk) 16:51, 3 September 2010 (UTC)
I don't have any objection to asking people to avoid "the 1800s", since that's consistent with some style guidelines, but a couple of years ago, we did a fairly energetic Google search and couldn't actually come up with a hit where one of the centuries ... I think it was "1700s" ... was used to mean 1700-1709, outside of Wikipedia. - Dank (push to talk) 03:08, 12 September 2010 (UTC)

This does keep coming up over[4] and over[5] and over[6] again, and it seems like everyone is loathe to actually provide guidance to editors on what to do the first decade of a century. I, myself, have been caught in a debate over the topic, in large part because I looked at WP:DECADE and WP:CENTURY and came up with a different answer than it seems most people reach. Based on the number of talk-page hits for "decade"[7], it seems I'm not alone. Instruction creep or no, we need a rule here to avoid confusion and ruffled feathers. How should one reword a sentence like "By the early 2000s, x happened" when one means the first few years of the decade spanning 2000 to 2009? If the answer is "it's okay if it's clear in context," I think that WP:CENTURY needs to say that. // ⌘macwhiz (talk) 21:26, 18 September 2010 (UTC)

  • The problem is that such text is often 'clear in context' only to the writer of the text and to no one else. This leaves all the readers having to guess what the author(s) may have meant. So I assert that '2000s' (for example) should never be used as it is inherently ambiguous and impossible to salvage. The 21st century is the correct substitute for century, but what is the correct text for the decade? Could it be be '2000s decade'? Short but awkward. Could it be 'first decade of the 21st century'? Completely correct, but long. What else? Something better? Hmains (talk) 02:27, 20 September 2010 (UTC)
Of those two options, I'd go with "first decade of the 21st century" despite its length; the phrase "2000s decade" just seems far too clunky, and I can't say I've ever seen it used in this fashion before... and I'm an avid reader. A sentence like "In the early 2000s decade, people dressed funny" just reads like it has this huge brick in it, right before the comma—it jars you out of your reading, because it's so unexpected. If this were a Neal Stephenson novel and not Wikipedia, I'd expect unusual linguistic constructs... Perhaps "decade of the 2000s" is a little better, but I'm not sure it flows better than the long phrase. Anyway, I'll concede your point about the possible confusion with "2000s", but if there's consensus on that point, the MOS needs to be updated to clarify all the places where such constructs are recommended so that the MOS isn't inherently ambiguous on the topic. We just need to establish what the best practice should be, and document it. // ⌘macwhiz (talk) 03:36, 20 September 2010 (UTC)

A question about SMS Goeben: "A total of thirteen men were killed and three were wounded." 13 and three? 13 and 3? Also, the addition I made to WP:MOSNUM to disambiguate, "in the article", that is mentioned up at #Currencies was apparently reverted ... anyone know why? - Dank (push to talk) 03:19, 12 September 2010 (UTC)

The reversion was this edit, but the edit summary doesn't seem to be connected to the reason for the edit. I guess we're talking about what the meaning of "is" is :) If we say something exists, does that automatically mean "in the article in question" without having to specify that? - Dank (push to talk) 03:33, 12 September 2010 (UTC)
I reverted the edit due to the disagreements noted above at #Currencies. If the article wasn't connected to the US, then the context isn't clear that US$ is being used, so it should be specified. It only needs to be specified once though. McLerristarr / Mclay1 03:56, 12 September 2010 (UTC)
My edit was: "in the article". Not really connected to that disagreement. - Dank (push to talk) 03:58, 12 September 2010 (UTC)
Yes it is. My disagreement is that the currency should be specified the first time it is used if it uses the same symbol as another currency even if that currency is not used in the article. McLerristarr / Mclay1 04:21, 12 September 2010 (UTC)
How about the first question ... thirteen and three, 13 and three, or 13 and 3? - Dank (push to talk) 04:30, 12 September 2010 (UTC)
The second point of the section states "Comparable quantities should be all spelled out or all figures: we may write either 5 cats and 32 dogs or five cats and thirty-two dogs, not five cats and 32 dogs." So 13 and three is wrong. The other two are both acceptable. McLerristarr / Mclay1 04:38, 12 September 2010 (UTC)

Anti-MOSNUM (actually, anti-arithmetic) sentiment alive and kicking

It's like 2 × 2 = 4.1. You just can't argue it's wrong in some venues. Tony (talk) 12:57, 12 September 2010 (UTC)

Tony, you mentioned precisely one organisation which consistently uses "percentage points" to refer to electoral swings: the Sydney Morning Herald. I pointed out to you that the Victorian Electoral Commission does the same. That's fine, no one is arguing that you, the VEC or the SMH are wrong. The trouble is, other outlets and commentators including the ABC, BBC and Australian Electoral Commission frequently (although perhaps not consistently) use "per cent swing". Guess what, neither is wrong! You can argue it's wrong in whatever venue you like, and you have done so loudly, frequently and occasionally quite rudely (although you have been subject to considerable rudeness and unnecessary reversion yourself). When others disagree with you, you have ignored most of their arguments and references and repeated: you're wrong! WP:MOSNUM! Sydney Morning Herald!, and then come running back here to make snarky comments about their "anti-arithmetic" stance. As I suggested, could WP:MOSNUM be clarified to refer to an electoral situation as well as a change in rates? Can you reach consensus here? If not, you can provide a reference to just one style guide which mentions this in an electoral context rather than as a change in interest rates (preferably confirming that "per cent swing" is incorrect)? To clarify, I have no problem at all with you changing "per cent swing" to "percentage point swing" in any and all election articles. What I do have a problem with is that you frequently use the term "point swing" with no context, and that you then feel the need to post on talk pages about how wrong everyone is. --Canley (talk) 07:12, 13 September 2010 (UTC)
Also, 2.0 x 2.0 = 4.1 if you round 2.025 to one decimal place. :) --Canley (talk) 07:15, 13 September 2010 (UTC)
It is much preferable and simpler to use "point" for swings, so that editors don't have to think through whether in each context it is ambiguous. But a compromise would be to accept the push to express swings (and margins) as percentages where unambiguous, provided "point" is used in other contexts. Tony (talk) 08:09, 13 September 2010 (UTC)

Ordinals cleanup

Following the deletions of {{st}}, {{nd}}, {{rd}} and {{th}}, I've been looking through other misuse of superscripting in ordinals. this search suggests that over 10,000 articles are presently using "1st" for "1st", which is contrary to the general discouraging of superscripting present in the MoS for a couple of years. Any suggestions on the best way to move forward with correcting this? Chris Cunningham (user:thumperward: not at work) - talk 09:04, 16 September 2010 (UTC)

That edit (looking for "nd", "rd" and "th" as well as "st") has been in my AWB selections without getting objections. I don't remember false positives, so I presume a bot could do it. Is that what you're asking? Art LaPella (talk) 20:23, 16 September 2010 (UTC)
If you write up a feature request for AWB I can add it to AWB genfixes. The effort from you is to consider/identify any exceptions: any templates/specific formats where superscripted ordinals are needed etc.? Rjwilmsi 21:30, 18 September 2010 (UTC)

Metric without decimals?

Use words for simple fractions; but use the numerical fraction form in a percentage, with an abbreviated unit, or mixed with whole numbers (an eighth of a millimetre, but 18 mm; 1+14 slices). Use figures for decimal fractions (0.025).

An eighth of a millimetre? Really? I won't bother to link to the brochure, but the SI explicitly says (this information may be out out of date) the fractions should never be used to represent the values of SI units.
This is almost promoting incorrect use (or representation, should I say) of metric units, it needs to be changed, in my opinion and there is truthfully little to argue against here for the ones who like picking fights. --Γιάννης Α. | 23:09, 20 September 2010 (UTC)

We're not bound by, nor are we bound to represent, SI usage. Saying "about 0.125 of a [metric unit]" looks far more precise, and misleadingly so, than "about an eighth of a [metric unit]" if one is just giving a rough average (say, of an animal's length) or an uncalibrated visual estimate (say, of perceived height).
One-eighth is a single significant digit; 0.1 gives significantly less information than 1/8, while 0.12, 0.13 and 0.125 give too many significant digits.
I was pretty hopeless at science lab and lab reports, but I suppose the way to express the idea in pure decimals that something was greater than a ninth of a [metric unit], but less than a seventh, would be 0.125 +0.018/-0.014 [metric unit abbr]. If anyone tried to impose that as a Wikipedia standard, I doubt they'd have much success. —— Shakescene (talk) 06:00, 21 September 2010 (UTC)
That makes very little sense. They're exactly the same value they carry exactly the same precision when you modify them with about. And we are bound to SI rules when using SI units. --Walter Görlitz (talk) 06:05, 21 September 2010 (UTC)
I would like to see where they found a meter stick in which the centimeters are divided into 8ths. ←Baseball Bugs What's up, Doc? carrots06:15, 21 September 2010 (UTC)
(edit conflict) Do you really say "that house is about 0.5 kilometre away" or "it seems to weigh about 0.3 kilo"? Most of the articles in Wikipedia don't cover scientific or technical subjects, but many of them are written by non-Americans or cover topics closely-related to metricated countries (e.g. French history), so their common, ordinary, everyday units will be metric. That shouldn't mean they have to follow all the rules of SI. —— Shakescene (talk) 06:24, 21 September 2010 (UTC)
No. We would say that house is about 500 meters away or it seems to weigh about 300 grams. --Walter Görlitz (talk) 06:40, 21 September 2010 (UTC)
If it's actually a guesstimate I'd usually say "half a kilometre" and "three hectograms", but then hectograms are fairly common in Italy but very rare elsewhere. A. di M. (talk) 16:53, 21 September 2010 (UTC)

Fractions are compatible with SI. In fact, two of the seven SI base units (metre and kelvin) are defined as fractions. One of the definitions even uses the word 'fraction'. See the official SI website: www.bipm.org/en/si/base_units.

  • "The metre is the length of the path travelled by light in vacuum during a time interval of 1/299 792 458 of a second."
  • "The kelvin, unit of thermodynamic temperature, is the fraction 1/273.16 of the thermodynamic temperature of the triple point of water."

Regards Lightmouse (talk) 09:02, 21 September 2010 (UTC)

  • My experience with metric use is that fractions (which for purposes of this paragraph excludes decimals) may be used when it is an exact definition or exact nominal value (for example, if 1 kilogram of material is to be divided as equally as practical into three parts). Metric fractions are also used during calculations. But the results of calculations are normally stated as decimals. Since encyclopedias normally present results, and not intermediate calculations, I don't think we should be presenting metric fractions in our style guide. They shouldn't be forbidden, but they shouldn't be encouraged either. Jc3s5h (talk) 16:23, 21 September 2010 (UTC)
  • The fact is, there's nothing wrong with e.g. 8+14 cm, but in practice such things are very rare. In most of the circumstances an American would write 3+14 in, an European would write 8.3 cm. (In guesstimates, you do often hear things like two kilometres and a half, but guesstimates usually use spelled-out words rather than digits, and anyway metric users are far more likely to estimate something to within a tenth than to within a sixteenth or an eighth.) Hence, I agree with Jc3s5h. A. di M. (talk) 16:51, 21 September 2010 (UTC)
  • As a practical matter, how do you sense the difference between .6 and .7 KG, for example? ←Baseball Bugs What's up, Doc? carrots17:05, 21 September 2010 (UTC)
    I'd say "it's about six or seven hundred grams" (or "six or seven hectograms" if in Italy), if I understand your question correctly. Saying "two thirds of a kilo" would sound terribly weird to me (unless it is something which is designed to weigh 666.6666667 g to within as good precision as practically possible, no examples of which spring to mind right now). A. di M. (talk) 17:35, 21 September 2010 (UTC)
  • I have no problem with discouraging things like 8 1/4 mm when 8.25 mm will serve as well without significant distortion; I was just thinking of simple fractions of a whole, like a third of a kilo or an eighth of a millimetre. But we don't really need to prescribe so exactly anyway because the specific point the MoS example is making (about using words with words and numerical fractions with abbreviations and mixed numbers) is just as clear with Imperial/U.S. customary units:

    Use words for simple fractions; but use the numerical fraction form in a percentage, with an abbreviated unit, or mixed with whole numbers (an eighth of an inch, but 18 in; 1+14 slices). Use figures for decimal fractions (0.025).

    —— Shakescene (talk) 18:31, 21 September 2010 (UTC)
The text "between .6 and .7 KG," should have been written " between 0.6 and 7.0 kg". There should a leading zero before the decimal point, there should have been a non-breaking space between the quantity and the units and that in this instance the units should have been in lower case. The case of the prefix can change the meaning of the text - for example, "1 mW" and "1 MW" mean two totally different things. Likewise the case of the symbol can change things dramatically - "t" is the symbol for "tonne" and "T" is the symbol for "tesla", similarly "h" is the symbol for "hour" and "H" is the symbol for "henry". Martinvl (talk) 20:15, 21 September 2010 (UTC)
Good observations, everyone. At least in Greece and England (let me explain), I've never seen fractions being used instead of decimals for providing metric measurements. Perhaps I didn't argue my case well enough.
Like the (Canadian living in Germany?) and (Italian-American?) confirm, in common usage (which WP tends to prefer), units are always written with decimals. Informally, one would say half a kilometre or 500 metres (either is used in Greece), but formally one would of course say "500 metres". This would always from my experience be written in decimals or converted to a division of it without a decimal separator between it, like 0,5 km to 500 m. Same goes for all metric units. Quarters and further divisions (as well as thirds) are almost unheard of, the typical metric unit being written as (e.g.) 330 ml—which is also used in England when using units such as these if a word about the object itself such as "can" in this case is not used.
I am sort of typing this in a rush, so please forgive any inconsistencies. But yes, like User:Shakescene pointed out, although I disagree with him on promoting usage of metric with fractions (one of the things the system set out to generally replace originally), contributors should be very, very careful in writing things incorrectly in the MOS as even simple examples like this can mislead people.
Finally, I'd like to extend this discussion into a debate (as I know I'm clearly not important enough to make this change—the community is) over if we should inform people in the style guide to avoid fractions when using metric units. So, do we have a consensus in favour of changing the example, at least (let this issue be resolved first)? --Γιάννης Α. | 17:59, 22 September 2010 (UTC)

I think it's a bit too strong to say that every use of fractions with metric units should be avoided. When people go to the market or to a butcher's shop, they often buy a pound or half a pound of something, the pound being a metric pound defined as exactly half a kilogramme. But thinking in metric pounds is very slowly dying out, and younger people are more likely to buy half a kilogramme or a quarter kilogramme than older people. Similarly for distances. You would never see something like "1/2 kilometre" on a street sign in Germany or France, but it's what people say. IMO the important points are the following:

  • Normally, only the simplest fractions are used with metric units (1/2, 1/4, 3/4 is probably a complete list).
  • Even these are only used in a lower register of speech. In particular, the contexts in which they are appropriate in an encyclopedia are very limited: In literal quotations, and when trying to connect a dry subject with the readers' experiences from daily life.

Fractions with metric units are usually inappropriate here, and that should be clear. But if we aren't careful with the formulation someone will go around and mechanically remove them even in the few cases where they are appropriate and better than the alternative. Hans Adler 00:20, 23 September 2010 (UTC)

MOS guidance is:

  • "Use words for simple fractions; but use the numerical fraction form in a percentage, with an abbreviated unit, or mixed with whole numbers (an eighth of a millimetre, but 1⁄8 mm; 1 1⁄4 slices). Use figures for decimal fractions (0.025)."

MOSNUM guidance is:

  • "Unless there is sound reason to the contrary, fractional parts of metric units should be expressed as decimal fractions (5.25 mm), not vulgar fractions (5 1/4 mm). However imperial units may use either form – both (5.25 inches) and (5 1/4 inches) are acceptable, provided that there is consistency in the way that the fractions are represented."

If you search for the word 'fraction', you'll find multiple references to fractions in MOSNUM. Some of which may be redundant. I don't know if MOS needs bringing into line with MOSNUM or vice versa. I think it's a guideline looking for a problem that doesn't exist. Editors appear to use decimal fractions for metric units anyway. The guidance could be removed and it wouldn't make the slightest difference to editor action. Lightmouse (talk) 08:58, 23 September 2010 (UTC)

  • Thanks, LM. A few points:
  1. I'm quite willing to consider the removal altogether of this guideline, but Hans is concerned to forestall an editor who might run around changing every occurrence of one into the other, without need. Is there any danger of this if the guideline is removed?
  2. MoS and MOSNUM must be harmonised on this matter (and all others, IMO).
  3. If a guideline is retained, I'd go for something much briefer. Do we have to teach people about the term "vulgar"? It would be so much easier to absorb if the verbiage were chopped out and bare examples used, such as:
And I've just done a double-take at this, in a current FAC: "Jewish Community Center 13-mile (0.54 km) away". Surely not: apart from the huge rendering of the fraction, the conversion is inconsistent with it. Tony (talk) 16:15, 23 September 2010 (UTC)
I have just aligned MOS with MOSNUM - the section on eigths of millimetre is now quarter of a pound. There has been no loss of meaning in MOS, while at the same time avoiding the contraversy surrounding vulgar fractions and SI units.
A journey of a thousand miles begins with a single step. Art LaPella (talk) 23:10, 23 September 2010 (UTC)

For proper disambiguation in date ranges between centuries I boldly edited from:
"The full closing year is acceptable, but abbreviating it to a single digit (1881–6) or three digits (1881–886) is not."
TO:
"In such case the full closing year is given, as abbreviating it to a single digit (1881–6) or two digits 1881–86) or three digits (1881–986) are not acceptable."—Iknow23 (talk) 06:44, 23 September 2010 (UTC)
Also, the prior one had a obvious error as the material was about (1881–1986), thus the three digit closing would be 986, not 886. —Iknow23 (talk) 06:49, 23 September 2010 (UTC)

I have to say, I like the pre-existing wording better. Why not just fix the obvious error? Tony (talk) 09:22, 23 September 2010 (UTC)
I can't see why we list any unacceptable examples in this case. Why not just:
  • A closing CE or AD year is normally written with two digits (1881–86) unless it is in a different century from that of the opening year, in which case the full closing year is given (1881–1986).
? PL290 (talk) 10:35, 23 September 2010 (UTC)
I SUPPORT PL's suggestion. Since unacceptable items were being listed, I wanted to add the missing one. But it does seem quite awkward that way. PL's is succinct and less prone to misunderstanding IMO.—Iknow23 (talk) 03:53, 24 September 2010 (UTC)
Absolutely, PL's wording express my feeling about how the dates should be. — Legolas (talk2me) 04:22, 24 September 2010 (UTC)
PL's is very good. But Iknow, I don't understand: you've switched from deprecating the two-digit closing range to supporting it (in accordance with the long-standing guidance). Is that correct? Tony (talk) 07:08, 24 September 2010 (UTC)
I agree with the above, but disagree about removing the part about "1996-1998" or whatever being acceptable [8]. It still is. To me, "1996-98" should technically mean year 1996 to year 98, though I realize it is common to actually write it like that. I mean who decided that that meant randomly keep the first two digits, but replace the last two with the ones after the dash? Grk1011/Stephen (talk) 12:52, 24 September 2010 (UTC)
  • Postscript: it's odd to me that the place where the two-digit closing range provides the greatest advantage is in infoboxes (space-poor); yet this is where I find I have to do a lot of changing from XXXX–XXXX. Tony (talk) 12:58, 24 September 2010 (UTC)

Tony, I was wanting to take this as a two-step process. First I wanted to clear up what it is currently saying. Then I was planning to propose 'deprecating the two-digit closing range' in all cases. I didn't want to overscope this present discussion.—Iknow23 (talk) 21:29, 24 September 2010 (UTC)
Grk, THAT is how I was understanding it in that "The full closing year is acceptable" in all cases, but it was pointed out to me that it was in context ONLY with the 'different century' senario.—Iknow23 (talk) 21:35, 24 September 2010 (UTC)

I think that was meant as an opinion, not as a clarification. I see no issue with with full closing year. Grk1011/Stephen (talk) 00:07, 25 September 2010 (UTC)
Ok, since we are going to go ahead and discuss it here...I agree that the full closing year should be acceptable in all cases. IMO "2008-10" looks unprofessional and I think of it a bit as like calling people by their first name instead of their last in articles. How do outside professional writing sources consider this issue?—Iknow23 (talk) 01:26, 25 September 2010 (UTC)
I think that 2010–2011 and 2010–11 should be both acceptable when the end year is in the same century as the start year, which IIRC is what the guideline had always said until short ago. A. di M. (talk) 04:54, 25 September 2010 (UTC)
Outside sources don't have infoboxes, and often not tables. I believe there should be encouragement to use just the two digits in these cases at least. They are space-poor. We do the same elsewhere for space-poor contexts (% rather than spelling it out, for example; same for numbers). PS Iknow, please note that an en dash is used for ranges, not a hyphen. This would allow us to judge the number-of-digits issue in isolation, in your example. Tony (talk) 07:42, 25 September 2010 (UTC)
I agree full closing year is also acceptable, though it depends on context in some hard-to-define way. In some cases, repeating the full information is cumbersome (as with page numbering), hence I left the existing "normally" in my original suggestion. Taking into account the futher comments, I now suggest:
  • A closing CE or AD year is normally abbreviated to two digits (1881–86), but may also be written in full (1881–1886). Where space is limited, such as in tables and infoboxes, the abbreviated form is used. Wherever a closing year is in a different century from that of the opening year, the full form is used.
PL290 (talk) 09:05, 25 September 2010 (UTC)
Sounds good to me. Grk1011/Stephen (talk) 13:28, 25 September 2010 (UTC)

From Wikipedia:Manual_of_Style_(dates_and_numbers)#Numbers_as_figures_or_words:

"Numbers that begin a sentence are spelled out, since using figures risks the period being read as a decimal point or abbreviation mark"

This looks to me like a hangover from the days of typewriters as it's a hugely implausible risk. There are already too many exceptions in this section and I suggest just removing this one. GDallimore (Talk) 15:03, 25 September 2010 (UTC)

They are talking about numbers, not years. Like it's still odd to start a sentence with "6", we always write it out. Grk1011/Stephen (talk) 15:44, 25 September 2010 (UTC)
Ok, I can accept the space limited usage of the abbreviated form. But I would like to remove the 'normally' from the same century usage as it (to me) connotates preferred. I would like equal weight given to both, such as:
  • A closing CE or AD year may be abbreviated to two digits (1881–86), or written in full (1881–1886). Where space is limited, such as in tables and infoboxes, the abbreviated form is used. However, when a closing year is in a different century from that of the opening year, the full form must be used.
Iknow23 (talk) 16:01, 25 September 2010 (UTC)
Upon further reflection, is the space limited sentence required. It sounds reasonable and all, but is two more digits really that much of an overflow? I'm not a programmer, so I'm just asking.—Iknow23 (talk) 16:07, 25 September 2010 (UTC)
It's not to do with programming, but with appearance. Let me diff you to two examples, tomorrow; I'll need time to find a suitable one. Tony (talk) 16:15, 25 September 2010 (UTC)
Hmmm. It seems odd to say that a 'full form' is sometimes not acceptable. That is the true information. An abbreviation is just an understood form of the original.—Iknow23 (talk) 04:26, 27 September 2010 (UTC)

I would just like to point out that depricating two-digit years would violate ISO 8601, which would interpret 2008–10 as the tenth month in the year 2008. Using the dash in any case would not comply as ISO 8601 requires a solidus (as in 2008/2010) to represent time intervals, but perhaps two-digit years were discouraged at least to avoid ambiguity. sroc (talk) 13:13, 12 October 2010 (UTC)

No, ISO 8601 uses hyphens, not dashes, so October 2008 would be 2008-10. Anyway, ISO 8601 doesn't apply to dates in ordinary natural-language prose. A. di M. (talk) 18:50, 12 October 2010 (UTC)
WP:ISO 8601 doesn't apply - at all - even in the ISO 8601 article! I would suggest
  • A closing year is written in full (1881–1886) or, optionally, abbreviated to two digits (1881–86), if in the same century AD (CE).
Looses a few words, un-ambiguous I think - the choice, where there is one, is left to the editor. The only minor quibble is that normal lifetimes, or periods of service are always unambiguous in DDDD-DD form - thus 1898-04 in a table of mayoralities seems to me un-exceptional, and not worth forcing the other dates 1894-98 etc to full form for. Rich Farmbrough, 09:58, 13 October 2010 (UTC).

BCE/CE vs BC/AD

I have been told that I'm not allowed to change BCE to the less ambiguously defined BC because of this "manual of style". Who can I go to if I want to discuss that rule?--Axiomtalk 22:44, 26 September 2010 (UTC)

There's nothing ambiguous about BCE. As far as discussing it, here might be a good place. ←Baseball Bugs What's up, Doc? carrots22:46, 26 September 2010 (UTC)
For whatever it's worth, I'd be against changing the rule. Both systems are widely used, both are taught in schools, and most readers understand both, so there's no real reason to restrict one of them, unless Wikipedia is taking the position that only the Christian system should be used. Since Wikipedia isn't an exclusively Christian encyclopedia, that doesn't seem useful or necessary. -FisherQueen (talk · contribs) 22:50, 26 September 2010 (UTC)
I concur with FisherQueen, there is no need to change the established system we use, least not for this user religious views. This users stated problem seems to be " I have a problem with inconsistancy, and a bigger one with using Secular-Academic terms like BCE/CE, which are arbitrarily invented to move away from traditional AD/BC"[9], which to me isn't a valid reason.Heiro 23:12, 26 September 2010 (UTC)
Wikipedia is available throughout the world, and imposing a specific religious viewpoint on wikipedia is, to say the least, not appropriate. ←Baseball Bugs What's up, Doc? carrots23:38, 26 September 2010 (UTC)
How does BCE make it not a religious viewpoint? BC/AD is a religious system centered around (though not on) the birth of Jesus Christ. BCE/CE is the exact same system with a different name. If you were truly serious about not using a Christian system of numbers, you wouldn't use the Christian system of numbers. (Long story short: I find the whole argument about 'i don't want to use a religious system' bollocks. It's religious. Changing the letters doesn't change that.) --Golbez (talk) 00:20, 27 September 2010 (UTC)
BCE and CE are non-religious. They use the same "Common Era" numbering system, without imposing Christianity directly on it. ←Baseball Bugs What's up, Doc? carrots00:24, 27 September 2010 (UTC)
The numbers were created by and for Christianity, though. The letters you use don't change that fact. If someone is so disgusted about Christianity being "imposed" upon them by the letters used, they should have the honesty to discard the numbering system as well. --Golbez (talk) 00:34, 27 September 2010 (UTC)
Originally, yes, but now it's become the standard and is no longer tied to Christianity as such. Also note that the original meter was based on the distance from the north pole to the equator, but that doesn't mean it still is. ←Baseball Bugs What's up, Doc? carrots02:55, 27 September 2010 (UTC)
Note here[10] that this religion-neutral construct has been around for nearly 200 years. ←Baseball Bugs What's up, Doc? carrots03:11, 27 September 2010 (UTC)
Golbez, you seem to be advocating discarding both AD/BC and CE/BCE. Is that correct? If so, what do you propose replacing them with? If not, do you agree with the OP that Wikipedia needs to change the Manual of Style to say that BC/BCE is no longer permitted? -FisherQueen (talk · contribs) 00:42, 27 September 2010 (UTC)
I'm in favor of discarding neither; we are descriptive, not proscriptive. If we want to use BCE/CE in scientific articles because that's preferred, that's fine. Though, I'm not sure I understand the OP, as I thought editing solely to change from BCE to BC or vice versa was frowned upon. But let's not delude people into thinking they're escaping some great Christian hegemony by changing two letters. --Golbez (talk) 03:17, 27 September 2010 (UTC)
You're right, it is frowned upon. Seems to me we had a recent brouhaha over this issue. In any case, the OP makes it clear that he wants to get rid of BCE/CE in wikipedia, because he's pushing a specific religion. ←Baseball Bugs What's up, Doc? carrots03:31, 27 September 2010 (UTC)
Although I'd personally prefer BCE and CE to be used because I am, well, an Atheist and that is simply my view of the world but also because it doesn't discriminate against the easily offended of other religions, I think BCE and CE should be preferred to BC and AD in articles on history and articles about non-Christian civilisations and countries, etcetera.

The English Wikipedia budges up to the Muslims and Jews, among others, so for the rest for whom we don't need to use a specific calendar, I believe historical dates should be provided in BCE and CE. This, I think, should include articles on history before the countries in question converted or were proselytised into Christianity, like Greece and Korea, for example.
What do my fellow Wikipedians think about this logical idea? --Γιάννης Α. | 03:55, 27 September 2010 (UTC)
Yes, Golbez, editing solely to change from BCE to BC or vice versa is frowned upon- that's what the OP wants to do. My comment should be taken in that context- not that there's any need to escape a great Christian hegemony, but that the OP isn't going to be able to install one. -FisherQueen (talk · contribs) 10:36, 27 September 2010 (UTC)
I'd like to point out that being American and raised Christian, BC and AD are the terms that come naturally to me. If this were Christian-pedia, that would be fine. But this is a global website, with many non-Christian readers. So the religion-neutral BCE/CE usage is totally acceptable. As a practical matter, AD vs CE doesn't come up that often, since a year without qualification is assumed to be AD or CE. And to the one who said, "Why don't they use their own numbering system?" the answer is, they do, within the context of their religion - both the monthly calendar and the yearly numbering. Jews, Muslims, Hindus, the Chinese, for example, all have cultural and religious calendars unique to them. Russia switched over to the Gregorian calendar although the Eastern Orthodox, a Christian sect, continues to base its holidays on the Julian calendar. In the Jewish calendar, today is the 19th day of the month called Tishri in the year 5771. (Here's a handy converter:[11]) That dating makes perfect sense to a Jew but is a mystery to most everyone else. For public interaction, for business and civil and historical purposes, it makes more sense to use the "Common Era" numbering scheme, since "everyone else does". ←Baseball Bugs What's up, Doc? carrots12:09, 27 September 2010 (UTC)
The world is inconsistent on this and so should wikipedia. Changing the era of a page is possible but requires consensus which means proposing it on talk pages and requires a lot of effort. This means that changes will be few and edit wars far less likely. To change the current wording would just open a whole can of worms. IMO most people who remembers the edit wars that this used to provoke are going to want keep things as they are as what we have is less imperfect than the alternatives.Dejvid (talk) 12:43, 27 September 2010 (UTC)
I have to point out that by now, BC is pretty much obsolete, while BCE has never had higher popularity, even in the US. I have never once in four years seen BC in any scientific readings (which I read a lot of) while lots used BCE (also, BCE is not religiously biased). Therefore, my vote is to work to change all BC to BCE. AD and CE are so rarely used that I do not think it is worth it to change them. Yankeesrule3 (talk) 19:36, 9 October 2010 (UTC)
Outside scientific, technical, academic and religious publications, BCE is still very rarely seen in the U.S. I can't imagine a reporter or announcer for one of the major networks (or even the Public Broadcasting System or National Public Radio) saying "B.C.E." without explanation. Nor are there any popular film or book titles I can think of that use "B.C.E." It all depends on your context. While Wikipedia is consulted all the time on academic topics, its inherent limits are continuously stressed in academic and scholarly environments (as in "Don't even think of citing Wikipedia in your footnotes!") I think most of Wikipedia's readers are still Christian or non-religious pre-college students or general-interest readers who will be unfamiliar with BCE. —— Shakescene (talk) 22:44, 9 October 2010 (UTC)
BC is by no means obsolete in America - I would say it's the standard usage, at least among Christians (which America mostly is, one way or another). BCE is often used by non-Christians in America, there just aren't that many of them and the subject doesn't really come up that often in casual conversation. And there's a good reason why any teacher worth their salt would disallow wikipedia as a reference: Because anyone can edit it. Wikipedia itself acknowledges that Wikipedia is not a reliable source, even within Wikipedia articles. It serves as a summary, as a guideline, a springboard to other research. ←Baseball Bugs What's up, Doc? carrots23:04, 9 October 2010 (UTC)
A convention I've seen in an oldish maths book is to used signed integers, but IIRC negative ones are off by one: they use "+1" for AD 1, "0" for 1 BC, "−1" for 2 BC, and so on. (I'm not endorsing that, anyway...) A. di M. (talk) 01:40, 10 October 2010 (UTC)

To address the original question. If the article was created and past the stub staging using BCE/CE, then it should stay. If the article was changed to that after it was establish it should be discussed and the original format should be applied. This is the basic rule of the Manual of Style. BCE should not be changed to BC just for the fun of it or vice versa. --Walter Görlitz (talk) 04:14, 10 October 2010 (UTC)

  • IMO, both BCE and BC are recognized by most readers nowadays. But the term “BCE” is certainly relatively new and I find it to be especially novel sounding in the spoken word. Only two or so years ago, there was a “Who Wants To Be A Millionaire” show where Regis Philbin mentioned a date and used “BCE” (Bee See EEE). The roughly 30-year-old contestant responded “What’s that mean?” There is no question that BCE is a relatively new word. I have also noticed that BCE is less common in narrated documentaries; presumably because it tends to call attention to itself more so in that form.

    Even for those who know what it means, I think readers tend to have a *neuron-trigger* of “something new” when they encounter instances of BCE. Accordingly, for articles directed to a general-interest readership (not the scientific papers to which our articles may refer), I think the newer “BCE” tends to be distracting just because more of us notice the choice of wording and that distracts from the thought. I am keen to avoid any writing style that calls attention to itself so I tend to avoid “BCE” to make the article read as fluidly as possible.

    I know that those who like to use the new terminology probably find it more satisfying and will claim it is “natural” to them. But I honestly think that as writers, their minds are actually noticing the writing style and being pleased with the choices rather than not noticing the writing style and being absorbed in the thoughts being conveyed.

    Finally, I don’t think “ambiguous” has anything to do with it. If one is reading an article on Egyptian pyramids, “It was built in 3000 BC” is just as clear as “It was built in 3000 BCE”. As far as I know, the admonition against changing one form to another is to avoid edit warring. Greg L (talk) 16:48, 11 October 2010 (UTC)

    Good point, which I hadn't thought about before. On the Corpus of Contemporary American English there are 22 occurrences of "BCE" dating back to 2008, 2009 or 2010 all referring to the epoch, compared to 224 occurrences of "BC" in the same period 90 of which (if I counted correctly) refer to the epoch. Frequencies differing by a factor of four should be something to be taken in account, IMO. A. di M. (talk) 10:30, 12 October 2010 (UTC)
    Those statistics are only about American usage - what about the global situation? I can clearly remember being instructed in Archaeology classes at university in South Africa, back in the late 1980s, to use BCE/CE. BC/AD was declared to be only acceptable if the subject is specifically religious history from a Christian pov. I know this is anecdotal and original research but it is perhaps at least indicative of a trend towards BCE/CE beginning significantly earlier outside of the US (which according to some critics is at times almost a Christian theocracy). Roger (talk) 11:10, 12 October 2010 (UTC)
  • Roger, university instructors write scientific papers that are directed to an international scientific readership. Scientific journals have rigorous style requirements and there is zero wiggle room arguing with the editors. So, certainly an instructor is going to advise a room-full of 19-year-old archeology students to use BCE. There are a number of writing conventions used when a paper is directed to a scientific readership, like this: “Humankind in its modern form has been in existence for ca 250 000 years (R. Meyers et al.).” But for an encyclopedia directed to a general-interest readership, like “Neanderthal,” we don’t use thinspaces as a delimiter to the left of the decimal point, we don’t use the scientific-style parenthetical in-text attribution with its Latin crypto‑jargon, and it is best to write “approximately” instead of “ca” (still more Latin).

    Do you know the underlying reason for delimiting numbers with only the American-style commas? Seems rather chauvinistic, no? Because a given country in Europe might have three ways to delimit numbers and primary-school children are taught all three, including the American way (using commas). Americans are taught just one. So unless it is an especially scientific article (hopefully directed to a scientific readership), it is best to use commas for delimiting on en.Wikipedia because it causes the least confusion for its readership and allows articles to be read most quickly and naturally without brain (!) interruptions.

    The same goes for the “ca” in scientific papers. Again, we think about the target readership. While it is certainly fun to write articles that look like they belong in a scientific journal, writing “approximately” is better.

    The issue between “BC” and “BCE” is similar; though it is less so about “confusion” it is still about making articles read quickly and most naturally. For general-interest readerships I prefer BC as I have found through observation and experience that it is part of a writing style that least draws attention to itself. But this is just a generality—even for me. I agree with the use of BCE in a few rare cases, like “Dating Creation” because it covers how a variety of religions date creation. For that particular article, concerns over religious sensibilities trump awkwardness (assuming one has already recovered from the cognitive dissonance of how the entire universe was created precisely 2,196,528 days ago on October 23, 4004 BCE). That article is quite the exception. Greg L (talk) 21:53, 12 October 2010 (UTC)

  • Agree with the "that least draws attention to itself" sentiment. I'm convinced that "BC" and "AD" are more commonly known and used by the average person (over "BCE" and "CE"). Apologies, but I'm not that interested in what academia has to say on this issue as (on en.WP) we write for all of the English-speaking readers in the world (a surprising number of which don't come from academia). I would prefer to see "BC" and "AD" (where required) used in all WP articles.  HWV258.  21:18, 12 October 2010 (UTC)
  • Oh, well… Since you opened the door that wide, I’ll step in. Click on the above “2,196,528 days ago”-link. Notice what it uses? Indeed, BC is pretty common. For the most part, if it is just a matter of style, things go best when Wikipedia goes with the flow. Greg L (talk) 22:01, 12 October 2010 (UTC)
I'm inclined to agree "WP:Not a learned journal" (different, not worse or better), although some editors seem to think either a. That we are are learned journal b. Ought to be one. c. Ought to look like one (for reasons of credibility either with academia, who - rightly - are are unlikely to be impressed; or with the general public, in which case it could be argued we are conveying a false sense of reliability, contrary to Sue Gardner's recent words, referenced in the Signpost). Moreover BCE smacks of the same style of writing that has "Smith vacated his residence" instead of "Smith left his home" - while we are not simple:, I always imagine the text is being read by an fourteen-year-old intelligent but non-native English speaker. Every stumbling block - every un-necessary stumbling block, however small, is a Bad Thing™. Rich Farmbrough, 10:14, 13 October 2010 (UTC).

The religious objection to BC/AD comes from their spelled out versions, In the year of (the/Our) Lord and Before Christ; non-Christian recognize Jesus neither as Lord nor Christ. Those are dealt with by the BCE/CE distinction, even if setting the zero-year is the same. See Anno Domini. The massive number of non-Christian English speakers suggests this is a serious issue for English Wikipedia. There are also massive numbers of Chinese, Japanese, Korean, and Eastern European users educated with "Common Era" or "Western Era" equivalents as their norm.

At a minimum, we could agree to exclude BC/AD from articles about the history of non-Christian cultures, about non-Christian religions, and from technical articles in the sciences where BCE/CE are used.

Rich, can't we deal with readability issues with a template link?--Carwil (talk) 15:06, 23 October 2010 (UTC)

Migrating Chronological Items from WP:MOS

On the WP:MOS page I proposed reducing the section Chronological items to a short paragraph that would serve as an introduction to the corresponding section in this article, much as I did with Units of measure and as other did with Geographical names. I have identified four items, each of a few sentences in the WP:MOS article that do not appear in this article. Three are non-contraversial insofar as they do not contradict anything here, but one might need some discussion. Does anybody have any objection to me cutting and pasting the three non-contraversial items into this article without further discussion and to open up discussion about the fourth here? Martinvl (talk) 11:46, 7 October 2010 (UTC)

This is premature. The matter is under discussion at WT:MOS, where there's concern at lopping off important and much-needed advice to editors from a centralised location. The first job, before even thinking about rationalising MOSNUM and MOS, is to harmonise the sections that have diverged. Tony (talk) 14:23, 7 October 2010 (UTC)
Tony, my request was to do the harmonisation. Martinvl (talk) 15:36, 7 October 2010 (UTC)
Ah, it's quite a big job, and where there are inconsistencies, both versions and the proposed compromise or fix should be specified on both talk pages (or at least a link from one of them to the other). I did this for Currencies a month or two ago. Tony (talk) 15:48, 7 October 2010 (UTC)
I identified four items in WP:MOS that are not in WP:MOSNUM. Three are non-contraversial, so I have migrated them. I will open up a discussion on the fourth later. Once the issues in the fourth have been resolved, the calendar items in WP:MOS will be a subset of those in WP:MOSNUM. Martinvl (talk) 12:26, 13 October 2010 (UTC)

Stylistical adjectival parenthetical imperial-metrical hyphenational insertions

"Avoid inconsistent usage. Write a 600-metre (2,000 ft) hill with a 650-metre (2,100 ft) hill"

And yet the parentheticals are non-hyphenational. In the above example I would 2000-ft hill as "two thousand foot hill" and "(2000 ft) hill" as two thousand feet hill" which is clearly(?) wrong.

(Brief) Comments? Rich Farmbrough, 13:42, 10 October 2010 (UTC).

ISO says no hyphen if the symbol (abbreviation) is used. We have followed that for a long time. Tony (talk) 06:15, 11 October 2010 (UTC)
Tony is correct; that’s the convention observed in all good technical writing. Hyphens never separate a unit value and its unit symbol. The ISO seldom pulls new ideas out of their butt; they generally merely sanctify and publish best existing practices—in this case, writing practices and proper sentence structure and punctuation. And when the ISO does pull ideas out of their butt and tell the world “Adopt this new way,” the idea sometimes falls on flat on its face.

Meters can be units of length or they can be gauges one installs into a control panel. Feet can be units roughly a third of a meter long or something attached below one's ankle. Yards can be units or something one grows grass on. There can be bridges for cars, and there are footbridges. There can be foothills. Ergo, when a measure is being used as modifier, “We have 200-foot bridges” is perfectly distinct from “We have 200 footbridges”. When the measure is used as a modifier, the hyphen nails down meaning and avoids having the readers’ eyes double back because of brief confusion. And rigorously adhering to the hyphenation practice helps the eye to parse occurrences where the expression is not being used as a modifier, like “We have 200 yard sales.” This hyphenation practice holds true for unusual units that don’t look like something else, like “radian” because our minds are accustomed to the convention. So it’s “A two-radian arc segment.” The need to clarify disappears with unit symbols like “ft”, “m”, “yd”. In fact, none of the SI’s unit symbols look like words (although “mol” looks a bit like “mole”). And the common U.S. Customary unit symbols generally don’t look like words either.

To state the obvious: a parenthetical like (2000 ft) is not a modifier and is unambiguous. And, IMO, “2000-ft hill” and “2000 ft hill” are both examples of poor writing; use the spelled out unit of measure with the hyphen. Unit symbols should never be used in a modifier for regular prose. To ensure we are all on the same wavelength, “The building had a height of 200 feet” does not have the “200 feet” being used as a modifier and therefore does not get the hyphen. Greg L (talk) 14:17, 11 October 2010 (UTC)

I would largely endorse Greg's comments above, but I feel I must point out that while "meter" has those two meanings, "metre" has only one. Further, the US Customary units symbols include the "in", which has tremendous potential for causing confusion. It is almost always better to spend the extra two characters and write out "inch" in full.LeadSongDog come howl! 20:58, 12 October 2010 (UTC)
(*Phew*) I’m glad I remembered to write “generally don’t”. As for metre, I’m not so sure: “A short fire broke out in two electrical metre-boxes at a building of Shankar, Dhaka on Saturday afternoon and (from Canadian woodworking) In electrical terms this is normally measured in watts per second or kilowatts per hour ( KWH as seen on the electrical metre). I’ve seen the same thing happen with the word “gauge” vs. “gage.” In the U.S. differential pressure relative to atmospheric is “gage” pressure (pounds per square inch-gage, or psi-g) because the instruments are “gauges.” In the U.K., it tends to be the reverse because the instruments are pressure gages; thus, kilopascals-gauge or kPa-gauge.” But the trend isn’t perfectly consistent within a country or even within a particular industry (so I go with the practices of the largest world-wide manufacturers of pressure transducers). Greg L (talk) 22:16, 12 October 2010 (UTC)
Thanks for the comments, I found the BIPM(?) document which says (to summarise) "hyphens for adjectives, except when abbreviating units" - it even - almost apologetically - says that using them is the "normal rules of grammar". Not that we follow ISO as Mosnum itself proudly dictates. I also agree with the point about binding adjectival phrases where it avoids ambiguity (the puns are not necessary - 2 metre rules - but relevant) , although I note there is a Germanic tendency towards compound words - spaces to hyphens, hyphens to conjunction - that I find un-necessary, and in some cases undesirable. In the present case I am fairly decided that the correct way to go is (although I'm not sure it's important enough for the MOS) "hyphenate and spell out adjectival uses". Rich Farmbrough, 10:28, 13 October 2010 (UTC).
It depends. You'd prefer "10-inch widget" to "10 in widget" and so do I, but when you get to "7 TeV collision" vs "7-teraelectronvolt collision""250 km/h winds" vs "150-kilometre-per-hour winds" I'd go with the former. A. di M. (talk) 13:18, 13 October 2010 (UTC) (amended at 10:43, 14 October 2010 (UTC))
I absolutely agree A. di M. But note that such a sentence would (should) only be found in a highly scientifically oriented article here on Wikipedia. If such a unit of measure was to find itself being used in an article truly directed to a truly general-interest readership, it would be expanded to make it more accessible and revised in structure, such as this: “Fermilab was the first accelerator capable of collisions with particle energies of one trillion electron‑volts (1 TeV).” Clearly we are now goosing butterflies with regard to best writing practices. Far too many of our technically oriented articles today have the equivalent of “7 TeV collisions” on the very first occurrence in the lede, where the writer walked away thinking he had done his wikipedian duty of “building the web” and looking way-cool by using a *sciency* unit symbol that is blue so a reader must (*sigh*) and click the link just to learn how you pronounce its components. Greg L (talk) 13:44, 13 October 2010 (UTC)
I just meant to give an example of a very long unit name. I've replaced it. A. di M. (talk) 10:43, 14 October 2010 (UTC)
I agree with you that unit symbols that are common in daily life, like km/hr (and mph in the U.S.) are so terribly familiar to a general-interest readership that they don’t need to be spelled out. “Investigators concluded he was driving at over 130 mph before hitting the tree” looks natural and so does one with the measure used as a modifier: “The storm, packing 75 mph winds by late evening, now qualified as a hurricane.” I’m not sure how the AP and The New York Times handle this, but I’m quite sure I would sail right over such constructions where non-spelled-out unit symbols on the first occurrence—even when they were being used as a modifier. For one thing, the unit symbols in the two examples I used here are not only well recognized, they can’t be confused with other words; ergo zero reason to spell out the full unit of measure and add the hyphen for the modifier. Greg L (talk) 19:24, 15 October 2010 (UTC)
I think that would be true were not Wikipedia an international encyclop[a]edia (cf. the characterization of USA Today, a national newspaper which has no single home-town, as "News from Nowhere"). Mph or km² make perfect sense on first sight to a regular reader of, respectively, the Houston Chronicle or Le Monde, but may need explanation to those who normally read the other paper. (And blue links or Wiktionary links to units are very cumbersome to use for the 99.5%+ of our readers who aren't registered editors who've enabled both JavaScript and WP:navigation popups.) If, as an outsider, you're reading the Buenos Aires Herald, The Rand Daily Mail, Arab News, The Jerusalem Post, China Daily, The Times of India, or The New Zealand Herald, you know you're visiting someone else's home and expect to encounter some untranslated words and terms; that shouldn't be true of Wikipedia. —— Shakescene (talk) 20:34, 15 October 2010 (UTC)
Context and ‘understanding your readership and writing for them’ is all important. If the article is on something like Porsche 911, it is perfectly appropriate to not belabor the article with “miles per hour” and “kilometers per hour” on the first occurrence because the symbol is transparent and familiar to virtually all the readership coming there. “Porsche 911” is the first article I guessed would be observing this practice and, lo and behold, that proved true. You are welcome to wade in and suggest that those automotive writers don’t understand proper practices when writing for an automotive readership. Methinks it will be an uphill battle. I also, just now guessed “Autobahn” would also not encumber its article as you seem to suggest. I guessed right.

As for your examples mentioning “China Daily” and its “we are the world” thrust; there are scores of Wikipedias in languages other than English. We write for readers whose first language is English. If we wrote otherwise, our articles would look like Simple English.Wikipedia. Oh, BTW, that link to “Simple English Wikipedia” is to the “Autobahn” article. Even there, they skip spelling out such familiar units of measure because the readers are assumed to be literate and have seen a speed-limit sign if they are actually interested in the Autobahn. Greg L (talk) 20:09, 17 October 2010 (UTC)

Reference style

If an article uses citation in the form

J. Wales (31 July 2008). "Main Page". Retrieved 2008-08-01.

where the publication dates are consistently distinguished in format from accessdates, is this a "breach" of this MOS guideline for "consistency"? Is it necessary to convert all such references to

J. Wales (2008-07-31). "Main Page". Retrieved 2008-08-01.

Some developed and even featured articles have citations in the former form, and have had so for years with stability. The intro to this guideline says: "If an article has been stable in a given style, it should not be converted without a style-independent reason." Comments? Gimmetoo (talk) 03:20, 19 October 2010 (UTC)

I don't understand in what circumstances "2008-08-01" or "2008-08-01" would ever be used. They are only to be used in long lists and tables for conciseness and sorting, and constrained even then, is my understanding. (Because "2008-08-01" can be taken to mean "August 1, 2008" or "January 8, 2008".) Certainly to convert "31 July 2008" to "2008-07-31" in the case you describe would be an abomination. The person who is using "YYYY-MM-DD" for accessdates is making an error, although it's an error I wouldn't spend any time correcting. Since accessdates are technical convenience for readers and bots and are not really part of the ref, and are subject to change at any time, I wouldn't worry about making them match the real parts of the ref. This is my opinion. Herostratus (talk) 04:50, 19 October 2010 (UTC)
I don't know if this opinion is correct, but our guideline against bare URL's is at WP:LINK#Link titles. Art LaPella (talk) 05:06, 19 October 2010 (UTC)
Better to convert all dates to plain English. Mr Stephen (talk) 07:01, 19 October 2010 (UTC)
Strongly agree with Mr Stephen. Tony (talk) 09:24, 19 October 2010 (UTC)
I slightly disagree. I think YYYY-MM-DD dates are logical and easy to read, just not very human. I don't mind using that style for access dates, though I would object to its use in prose. I don't think that the citations need one consistent date style, but if they do then it should be something less robotic, like DD Month YYYY. (By the way, I've never seen anyone use YYYY-DD-MM; YYYY-MM-DD is completely standard (ISO standard, in fact), and there is no ambiguity, unlike for DD-MM-YYYY and MM-DD-YYYY dates.) Ozob (talk) 11:04, 19 October 2010 (UTC)

I agree with Tony & Stephen that

J. Wales (31 July 2008). "Main Page". Retrieved 1 August 2008.

is preferable to the example with both publication and accessdate in ISO. But that's not the question I'm asking. The question here is whether, in references, the publication and accessdate can be consistently distinguished by format, or whether they must have the same format. There are quite a few articles where the publication dates are in dmy or mdy, and the accessdates are in iso. As an aside, there is no ambiguity in the iso accessdate in the example, even for someone unfamiliar with iso; reading the accessdate as January 8 puts it before the publication date, which is obviously wrong. Gimmetoo (talk) 13:20, 19 October 2010 (UTC)

  • The question is related to the article Ursula Andress, where I was intending to make some accessibility improvements, but also spotted the mixture of 'dmy' and 'iso' formats, which I felt looked odd. As the accessibility changes can be controversial, I first outlined what I saw as problems at Talk:Ursula Andress#Accessibility and dates. The section WP:Manual of Style (dates and numbers)#Format consistency seems unambiguous to me (Dates in article references should all have the same format), but Gimme has pointed out that the present references have a kind of consistency, in that the publication dates are all dmy and the access dates all iso. Although I can see the logic in that, I still don't think that really fits with the principle of maximising consistency, and in particular the guidance at Format consistency. I think it would help both of us if others were able to help us reach a consensus (or at least settle the issue one way or the other!). The result obviously may have some bearing on whether the Format consistency may need some amendment, so this is probably the best place to ask. Thanks in advance for your insights. --RexxS (talk) 16:50, 19 October 2010 (UTC)
    • There is, in my opinion, no need to worry if, despite other dates on the page being in English, some, most or all "accessed" are in numeric. However I would not object to them being changed either. It is a moot point whether their display is actually useful for WP. Rich Farmbrough, 16:28, 20 October 2010 (UTC).

The presence of the format YYYY-MM-DD does not demonstrate that the publication in which it appears has adopted ISO 8601. In the absence of an explicit definition of what the format means in a particular publication, such dates are ambiguous.

Also, if one were to apply ISO 8601 to the publication date and one wanted to give the publication date of a current monthly magazine, one would be obliged to write 2010-10, which most uninitiated readers would not understand. Jc3s5h (talk) 16:48, 20 October 2010 (UTC)

  • As Rich says, there is no issue. It is generally accepted that the access date may be in YYYY-MM-DD format even when the publication date is spelled out. It is a legacy from when citation templates required such input, but it remains accepted practice. Some would like to see all dates spelled out in full at all times, but a recent (enough) RfC rejected such a proposal with specific regard to footnotes. In addition, we do not need another pointless discussion about some people's views on the ISO8601 and the YYYY-MM-DD date format. wjematherbigissue 19:01, 20 October 2010 (UTC)
  • If not for the unfortunate widespread use of shorthand for year-only date ranges, Jc3s5h, that last format would not pose a problem. ―cobaltcigs 03:28, 21 October 2010 (UTC)

Speed: 'km/h' versus 'kph', 'kmph', 'km/hr' 'kms/hr', 'KmH' etc

Note the following:

It has been discussed before.

Regards Lightmouse (talk) 16:26, 22 October 2010 (UTC)

In my experience, 98% of the times the symbol is km/h and 2% of the times is km/hr. I don't recall ever seeing kph except on this very talk page. (BTW I've downshifted the header levels in your quotation lest they mess up the lead and the ability to edit.) A. di M. (talk) 16:36, 22 October 2010 (UTC)
Maybe this is a nationality thing. I don't recall ever seeing km/h here in the UK - it's always kph (although naturally we usually use mph for speeds outside science!). And kph appears in the Oxford English Dictionary, whereas km/h does not. -- Necrothesp (talk) 17:01, 22 October 2010 (UTC)
Don't think so. Brits are still stuck on imperial, and without the divisor, so you are more likely to see mph and mpg than km/h. ;-) --Ohconfucius ¡digame! 03:55, 23 October 2010 (UTC)

If you have a car, take a look at the metric speed on the speedometer. What does it use? Lightmouse (talk) 17:18, 22 October 2010 (UTC)

In all eight Australian jurisdictions, it is "km/h". No ifs or buts. What does the ISO say about it? I note that Pfdpfd has seriously breached the civility code and appears to be unwilling to engage on the matter. Nevertheless, this is one to put in your table of issues that arise from time to time wrt units, That kind of organisation can prioritise what should be discussed, what should be avoided for the moment, what objections should be disregarded (this one, i'd say, with its theatricals. LM, can you establish such a table we can all consult? This would help to bring under control the large number of connondrums. connecting thebot runs with thethe authoritiesl=0000000000-----================--------------------------------------[[[[[[[[[[[[[[[[
A GMC American vehicle, model year 2010, has a speedometer labeled "km/h". Jc3s5h (talk) 17:31, 22 October 2010 (UTC)
"I note that Pfdpfd has seriously breached the civility code and appears to be unwilling to engage on the matter." - a) Yes, I did. b) Why do you "note that ... "? c) Appearances can be deceptive. d) What leads/led you to state that I appear to be unwilling? e) What is the point of raising the issue of lack of engagement? f) What is the point of discussing the issue of lack of engagement?
I agree with Lightmouse's suggestion: Wikipedia:A nice cup of tea and a sit down. Pdfpdf (talk) 09:54, 23 October 2010 (UTC)
In the UK, they tend to say "km/h" on the inner dial (and "mph" on the outer). Chances are there are plenty of people who simply haven't noticed that it's not "kph", since it's not a unit that gets used particularly often and "kph" is rather more common outside formal contexts. Pfainuk talk 17:35, 22 October 2010 (UTC)
Google Books says km/h, 5,220,000 hits; kph, 127,000 hits; km/hr, 960,000 hits; kmph, 20,900 hits; kms/hr, 11,600 hits. See also User:Art LaPella/Because the guideline says so. However, my practice is not to edit a specific page (assuming the objection is limited to a specific page) that gets someone that excited; there are 3 million other pages, and eventually such changes will look standard. Art LaPella (talk) 21:46, 22 October 2010 (UTC)

Isn't the real problem here en-masse conversions from one name of a unit to another because of some OCD compulsion about consistency? Does it really matter which one is more common? No. The meaning is clear either way. The policy on semi-automated edits is clear: Unless a change is completely non-controversial, it shouldn't be done in a semi-automated fashion. Gigs (talk) 00:54, 23 October 2010 (UTC)

"The policy" probably refers to the WP:AWB instruction "Don't do anything controversial with it." But Wikipedians argue about everything. So if an explicit guideline doesn't render an issue non-controversial, doesn't that mean AWB should be restricted to spelling? Art LaPella (talk) 05:17, 23 October 2010 (UTC)
I am not convinced the meaning is clear either way. How is a person who lives in a country that has used SI for a long time, and learned English as a second language, supposed to figure out kph if he/she never learned mph? The abbreviation kph does not contain the normal SI symbol for meter, nor does it contain the normal SI symbol for division. Jc3s5h (talk) 01:13, 23 October 2010 (UTC)
I'm properly fearful and hesitant to suggest this, but I have the feeling that given anglophone Wikipedia's diversity of readers and articles, and because it seems quite plausible that many might not recognize "kph" while others would be perplexed by "km/h", the best solution might be yet another custom-fitted Wikipedia convention that should be understood by most, i.e. "km/hr". If someone (or some well-controlled 'bot), after proper and due consensus had been achieved following sufficient outreach, were to change both the kph's and km/h's to km/hr, it could be argued that one wasn't imposing one group's fancies arbitrarily on the other, but rather making the unit of speed/velocity clear to all. —— Shakescene (talk) 05:09, 23 October 2010 (UTC)
I don't buy that there's anyone who could be perplexed by "km/h" but not by "km/hr". A. di M. (talk) 09:35, 23 October 2010 (UTC)

km/h is the scientific way of writing it. That or kmh-1. McLerristarr / Mclay1 08:56, 23 October 2010 (UTC)

It appears to me that there is no single univerally-used abbreviation.
So it appears to me that if one single abbreviation is universally used, whatever it is, there will be a notable portion of the population who won't agree with it.
Thus, changing direction, why is it necessary to have a single abbreviation used "universally"?
Why not just leave things as they are?
It appears obvious to me that if someone put kph in there, that's because they use kph. I don't see what classifying kph as "wrong" achieves.
My (much more polite) 2c worth. Pdfpdf (talk) 09:58, 23 October 2010 (UTC)

Many editors use non-standard, verbal, or specialist terms when a more accessible and/or standard term will do the same job. It was a surprise to me to find 'km/h' being questioned. Janitorial edits, by definition touch many articles so surprise questions are a common experience. I'd be grateful for mosnum to have guidance on this. I propose changing:

  • "When unit symbols are combined by division, use a slash to separate the symbols (e.g., for the metre per second use the symbol m/s, not mps)"

to

  • "When unit symbols are combined by division, use a slash to separate the symbols (e.g., for kilometre per hour use the symbol km/h, not kmph, kmh, or kph)"

Lightmouse (talk) 10:45, 23 October 2010 (UTC)

According to the SI (here), the abbreviation for "hour" is "h" and the abbreviation for "kilometre" is of course "km". In my opinion, it doesn't matter what some people may use, "km/h" is the only correct way of writing it, unless you use kmh-1, but that's too technical for common usage and would confuse some people. McLerristarr / Mclay1 10:51, 23 October 2010 (UTC)

When using metric units always follow the official SI standards absolutely strictly. No variation. No debate. Otherwise what is the point of even having any standards at all? Roger (talk) 11:07, 23 October 2010 (UTC)

As every computer scientist knows "The wonderful thing about standards is that there are so many to choose from." -- PBS (talk) 21:37, 27 October 2010 (UTC)
When in the June 2008 rewrite, we tackled this, and it was pretty much agreed that "km/h" was pretty much how to write things by default, with the possible exception of using "kph" in car-related articles for consistency with "mph". "km/hr" is just silly given "km/h", "kms/h" is wrong (symbols aren't pluralized), "kms/hr" is just horrible, "KmH" is plain wrong (kelvin metre henry), "kmph" isn't used at all (it's kph when written with the "p" for "per" instead of the /), so it leaves us with the correct SI version of km/h (default, vast majority of cases), and the possible occasional "kph" if circumstances warrants it. Personally I'd use "km/h" all the time since no one could possibly be confused by it, while "kph" will frustrate some readers. Headbomb {talk / contribs / physics / books} 11:56, 23 October 2010 (UTC)
Agree with Headbomb. The accepted abbreviation is km/h. Michael Glass (talk) 12:28, 23 October 2010 (UTC)
Dear Mr Glass. I, too, agree with Headbomb.
However, Headbomb is NOT saying "The accepted abbreviation is km/h". I suggest you might like to re-read what he wrote - he chose his words quite carefully.
Also: "The accepted[by whom?] ... " Pdfpdf (talk) 13:01, 23 October 2010 (UTC)
Agree with Headbomb and Mclay1. Headbomb does say "Personally I'd use "km/h" all the time since no one could possibly be confused by it, while "kph" will frustrate some readers." Sounds cut-and-dried to me. The analogy with "mph" I don't find convincing. Um ... pdfpdf, you were a little over the top on LM's talk page. Have you apologised there? Tony (talk) 13:05, 23 October 2010 (UTC)
Dear Tony1. Your edit comment says: km/h please, unless there's a good case put for kph
That is indeed what Headbomb said, and at the risk of repeating myself, I agree.
Note that your response here is not conveying the same message as your edit comment.
Note also that "The accepted abbreviation is km/h" does not convey the same message as "km/h please, unless there's a good case put for kph".
Regarding your concluding sentence and question, short answer: Yes I was. Yes I have apologised. Pdfpdf (talk) 13:25, 23 October 2010 (UTC)

"km/h unless there's a good reason for using kph"

Speedometer showing km/h.

I think we have reached a consensus: "km/h unless there's a good reason for using kph"

What do other people think?Pdfpdf (talk) 13:33, 23 October 2010 (UTC)

There's hope. Remember that this discussion could equally have focussed on any of the other synonyms such as 'KMH', 'KmpH'. It would be worth testing the good reason clause before we get too excited:
  • "Do we have any Wikipedia articles where a reasonable reader requires anything other than 'km/h'?
I can imagine that it's useful to mention synonyms in unit definition articles. Can anyone identify any others? Lightmouse (talk) 13:56, 23 October 2010 (UTC)
Remember that this discussion could equally have focussed on any of the other synonyms such as 'KMH', 'KmpH'
Actually, I disagree - for much the same reasons as those given by Headbomb. Goodnight. Pdfpdf (talk) 14:07, 23 October 2010 (UTC)
Agree. The rule should be to use km/h; if this rule prevents you from improving Wikipedia, ignore it (not that I can think of any case in which it does). A. di M. (talk) 16:20, 23 October 2010 (UTC)
I agree with A. di M., and I would add that in many instances, the quantity will be accompanied by a conversion to or from miles per hour, which should resolve any uncertainty on the part of readers who are not fluent in SI. Jc3s5h (talk) 16:41, 23 October 2010 (UTC)
I think it should be noted that the US Metric Association accepts km/h as correct and kph as wrong: http://lamar.colostate.edu/~hillger/correct.htm Australia also uses km/h www.comlaw.gov.au/comlaw/.../NationalMeasurementGuidelines1999.pdf Michael Glass (talk) 23:47, 23 October 2010 (UTC)
A di M: I'm a little confused. Are you saying that:
1) The rule should be to use km/h; if this rule prevents you from improving Wikipedia, ignore it
is similar to:
2) km/h unless there's a good reason for using kph?
Or are you saying they are different?
If different, what is the essence of the difference?
For example, is what you are saying similar to:
3) use km/h unless there's a good reason not to,
the essence being that kph may not be the only thing that may have a good reason?
(Notwithstanding the fact that I can't think of any case that may have a good reason.)
FWIW: My personal opinion is that 3) is better than 2), and as I interpret 1) to be very similar to 3), I therefore think that I agree with you. Pdfpdf (talk) 11:11, 24 October 2010 (UTC)
Yes, what I mean is essentially 3), but I also think that situations where there's a good reason to use something else are so rare that the possibility of exceptions needn't even be explicitly mentioned in the guideline, as it's already covered by WP:IAR. A. di M. (talk) 11:18, 24 October 2010 (UTC)
OK. Good! I'm happy. Pdfpdf (talk) 14:45, 24 October 2010 (UTC)

The phrase good reason is redundant (you may not be aware that wp:iar could be applied on the end of all mosnum guidance) and subjective (because we don't know what you think is a good reason). We've merely moved the debate from a main clause to a subclause. I see now that you (Pdfpdf) can't give an example of a Wikipedia article (other than unit definitions and quotes) where good reason applies. Would you mind looking at the articles you reverted and seeing if any of those shouldn't be changed to 'km/h'? In order to downgrade the temperature of such edits (in response to well-taken comments by people like Gigs and yourself), I'd like to have mosnum guidance that can be taken to BAG. Lightmouse (talk) 12:47, 24 October 2010 (UTC)

Hi Lightmouse. I'm less than keen to review the random selection I reverted whilst I was annoyed. From my side of the hill, my major issue is consultation and consensus. I think you are now aware that you should check which way the wind blows BEFORE letting loose with the bot. Whether they are furlongs-per-fortnight, cubic light-years, or square microns is secondary. If you want to change them to km/h, ask first, and no-one objects, I'm happy. Pdfpdf (talk) 14:45, 24 October 2010 (UTC)

As it applies to automotive-related articles, if (big ‘if’) “km/h” is most common for car manufacturers and current, reliable literature on the subject, then that is what Wikipedia should be doing for its automotive-related articles. That statement does not apply to the practices that are universally or near-universially observed in other disciplines, like ‘Rocket sled’. Greg L (talk) 17:24, 26 October 2010 (UTC)

Other compound units

The above discussion focused on km/h versus kph. But the automobile conventions as quoted above are in conflict with this page because they recommend rpm instead of rev/min, mpg instead of mi/gal, and mph instead of mi/h. I know only American usage, but I think rpm, mpg, and mph are all much more common than their alternatives. It seems to me that the current guideline needs an exception, something like: However, if reliable sources consistently use a different style, such as mph instead of mi/h, follow the sources instead. Ozob (talk) 14:01, 24 October 2010 (UTC)

Sorry Ozob, but I feel you have missed a few of the important points from the above. (e.g. the above discussion is more about thinking first and seeking consensus BEFORE making wholesale changes.) As for your other points: yes, I too was interested to see that the "standards" are rpm, mpg, mph and km/h - km/h does stand out as different from the others. But don't get me wrong - I'm not complaining, just noting that km/h is different. Pdfpdf (talk) 15:00, 24 October 2010 (UTC)
I disagree. We cannot have yet another "if the sources contradict international standards" policy. We already have this policy in the debate concerning the correct measurement system to use. The point is there is no good reason to use any other unit formatting as the one expressed in the international standard. There is not one reason to use "kph" instead of "km/h"; no good reason to use "rev/min" instead of "rpm". Here is another question: why is there a whole debate about this when it seems that everybody agrees except Pdfpdf? If the number of opposing people is largely inferior to the number of agreeing people, then why is there such a long debate? Xionbox (talk) 14:15, 24 October 2010 (UTC)
Actually Xionbox, I disagree with most of what you have said. You seem to have missed most of the important points of the above discussion. I assure you that "the point" has almost nothing to do with what you say, and please supply some reliable sources if you are going to make absolute statements like "there is not one reason" - such a statement is almost certainly inaccurate.
As to: "everybody agrees except Pdfpdf" - well, that is simply a false statement.
"why is there a whole debate about this" - For a number of reasons. I suggest you reread the conversation. Then the answers to your other questions will become obvious. Pdfpdf (talk) 15:00, 24 October 2010 (UTC)
There is already this provision in the MOS.

Some disciplines use units not approved by the BIPM, or write them differently from BIPM-prescribed format. When a clear majority of the sources relevant to those disciplines use such units, articles should follow this (e.g., using cc in automotive articles and not cm3). Such non-standard units are always linked on first use.

Headbomb {talk / contribs / physics / books} 14:16, 24 October 2010 (UTC)

If I understand what you mean, MOSNUM already says: "Exceptions include mph for the mile per hour, psi for pounds per square inch, etc." A. di M. (talk) 14:30, 24 October 2010 (UTC)
Precisely. Pdfpdf (talk) 15:03, 24 October 2010 (UTC)
Pdfpdf, you are right. I did in fact improperly read the debate and I sincerely apologize for my previous remarks. Concerning a good reason to use "kph" instead of "km/h", I suggest people participating to this debate to "please supply some reliable sources". In fact, I have visited and lived in several countries (some which use the SI and some that don't), and have only once seen "kph": in a paper of the university student who probably did not know of the international standard for writing SI units. Hence, and because "km/h" is an international standard, I think it's up to the people wanting the possible usage of "kph" to propose sources in which "kph" is more accurate to the context than "km/h". I have reread the current debate and have not seen any real source apart from an example: "rpm" instead of "rev/min", which is not a good example since not directly linked to the subject of the debate. Xionbox (talk) 23:23, 24 October 2010 (UTC)
Thanks for your reply. I doubt that you'll be surprised that I agree with you. Pdfpdf (talk) 10:11, 25 October 2010 (UTC)

My point—which I guess I failed to make clear above—is twofold. One, the "clear majority" provision is under "scientific and technical units", so it might be interpreted to not apply generally. Two, MOSNUM mentions mph and psi as exceptions to rule about division and then says "etc." It seems to me, judging from the conversation here, that the "clear majority" provision should apply more broadly, and that the "etc." should be taken to mean "clear majority". Currently MOSNUM doesn't do this. Does anyone else see this as a problem? Ozob (talk) 00:59, 25 October 2010 (UTC)

I'm not sure what strategy (if any) lead to the current structure. But I'll point out a few things about how scientific and technical units differ from other units. First, the only country where both (1)non-metric units are heavily used and (2) a great variety of websites, books, and government-sponsored calibration services are readily available is the US. Second, the US has never codified the customary weights and measures; there is no list one can point to and say "if it isn't on this list, it's illegal". Third, NIST is pro-SI and isn't interested in providing official definitions, or anything else, related to customary units. Fourth, SI does provide a definitive body of definitions and symbols. So the guideline that we stick to SI for scientific and technical units, unless there is a clear majority of pertinent sources supporting some other usage, makes sense, because there is a definitive standard from which to depart. In the case of customary units, there is no comparable standard to use as the default. Jc3s5h (talk) 03:01, 25 October 2010 (UTC)
In the goal to comply with everybody's demands, what do you think of the following proposal for the MOSNUM: For SI units, always use the exact formatting defined by the BIPM. For all other customary units, use whichever formatting is the most used by reliable sources. Xionbox (talk) 12:08, 25 October 2010 (UTC)
I think there are some SI units that are better described in formats different from those defined by the BIPM, such as "cc" for cubic centimetre in automotive contexts (where the SI would seem to be "cm3"). This is provided for by the guideline as is, but appears to be excluded by your proposal.
I'd also note in passing a point that we need to be cautious on. We need to be careful about what "a clear majority of pertinent sources" or "most used by reliable sources" means. For example, in measuring astronomical distances of the order ~1015 metres or more, Wikipedia practice should prefer the light year because it is easier to understand and preferred by popular scientific literature - even though the parsec is rather more common in academic work. (It should go without saying that, in this case, the SI measures of petametres, exametres, zettametres and yottametres are so rare that they shouldn't even be used in conversion.) Pfainuk talk 17:10, 26 October 2010 (UTC)
I agree with Pfainuk regarding "a clear majority of pertinent sources" or "most used by reliable sources". For large stellar distances, light years are preferred. If the subject is about the parallax distance measurement to nearby stars, the parsec would typically be best. For any given subject, it is best to follow the practices of modern, most-reliable literature on the subject in order to ensure our readership is best primed for further studies on the subject. Greg L (talk) 17:44, 26 October 2010 (UTC)
As for astronomical distances, I like to use light years with conversions to (kilo|mega)parsecs (in general-interest articles) or vice versa (in more advanced articles). A. di M. (talk) 18:04, 26 October 2010 (UTC)
Yes - rereading that I was perhaps a tad sweeping. Of course, in more technical articles, parsecs should go first. But I am certain that there are cases where the most reliable sources for a piece of data are academic and thus prefer parsecs, but where we are writing a less technical article and should put light years first. Pfainuk talk 21:12, 26 October 2010 (UTC)
My point is that if accessibility is the goal, the best way to achieve it is to give both units. Which one should go first is a relatively minor detail. A. di M. (talk) 00:24, 27 October 2010 (UTC)
I agree; give both. The important part is to ensure both the U.S. Customary and the parenthetical are real-world units of measure commonly used by current, most-reliable literature on the subject. Often—but not always—that will be a unit from the SI. At times, it might a bastardized variation of SI or outright circumvention of the SI such as hours or µgals, or (*sound of audience gasp*) cc (for Honda motorcycle engines) or kilocalories per hour (in the discipline of metabolic activity). We do not do our readers a favor by writing really *sciency*-looking stuff like the vigorous exercise was shown to increase metabolic activity by 350 J·s–1 when we could instead write the vigorous exercise was shown to increase metabolic activity by 300 dietary calories per hour. The latter is infinitely clearer to a general-interest readership because we simply go with the flow instead of try to lead the world to a new and brighter future where only one set of rules applies to all of Wikipedia’s articles equally. Greg L (talk) 01:14, 27 October 2010 (UTC)

SI unit 'micrometre/er'

The term 'micron' had an official status between 1879 and 1967. It was officially removed from SI 40 years or so ago. The terms 'centigrade' and 'micron' continue to be used by many people but I think it would be worth having mosnum guidance on the use of the both terms 'centigrade' and 'micron' within WP.

Here are some issues that occur to me:

Ability to guess the meaning
The benefits of SI are that they do away with special names that have to be learned or explained. It has a simple format (prefix plus unitname). From that simple format, you can guess the size and/or the unit. Somebody may not have heard the term microwatt before but they should be able to realise it's 'micro' plus 'watt'.

Accessibility - micrometre/er is universal, the term micron isn't
Scientists are aware of both terms and there is no domain or geographical region which doesn't use the SI term. The term 'micron' is used by some (the term 'some' meaning anything less than 100%) scientists in some technical domains and some geographical regions. However, some ordinary people don't know what a micron is (as is shown by the multiple cases where it has to be explained in text). Thus the SI term is more accessible.

Ambiguity The American spelling of micrometer is ambiguous as a single word. Metrology has several ambiguous words (e.g. 'foot', 'yard', 'minute', 'second', 'mole'). However, the meaning of micrometer is always clear from the text (e.g. "the wavelength is 54 micrometers", "we measured it with a micrometer"). That may apply with most of the ambiguous unit terms used in WP articles.

Similar terms as the scale changes e.g. nanometer, micrometer, millimeter It is convenient to be able to scale units using similar terms. This can be seen in text such as US NIST: "from deep violet at 400 nanometers (nm) to near infrared at 4 micrometers (μm)"

Guidance Style guides provide the benefit of consistency across the project. WP isn't tied to one particular domain or geographical region. As with guidance on several issues, we can come up with our own consistent guidance so editors don't have to debate the issue on each article. I propose the following guidance:

  • Use the term 'micrometer' or 'micrometre' rather than the term 'micron', except where required by quotes or text describing history of units.
  • Use the term 'degrees Celsius' rather than the term 'degrees centigrade', except where required by quotes or text describing history of units.

Comments welcome. Regards Lightmouse (talk) 17:13, 20 October 2010 (UTC)

I don't see the necessity of specifying this in the MOS/MOSNUM. But I'm not against a general "don't use obsolete unit names" covering micron/centigrade/fermi/etc... Headbomb {talk / contribs / physics / books} 18:18, 20 October 2010 (UTC)
This is a result of a dispute that began on User talk:Lightmouse where I objected to his changing the familiar word "micron" to the infrequently used term "micrometer." I am also much in favor of gravitating towards SI units whenever there isn't a particular reason (I can think of a few) for retaining a traditional unit in a particular discussion. However "micron" already is an SI unit inasmuch as it is precisely synonymous with micrometer. It is widely understood and employed and (I am quite sure) is much more frequently used than "micrometer" in every field which deals with such units of distance.
As a quick check, I did a search of 3992 emails I have received regarding an area of optics, especially infrared optics (where wavelength is very often mentioned) and only found 4 instances of "micrometer" being used whereas well over half of the emails did use the word "micron" for that purpose. Thus "micron" was favored by a ratio of 500:1. Having been in this field for >30 years, I haven't seen any trend towards the use of "micrometer" or any reason that the previous term would be abandoned.
It is true, as Lightmouse points out, that using standard SI units and prefixes allows one to immediately understand a new prefix-unit combination when it first appears. That is good. However to blindly impose that principle in the particular case of a very common term which is so widely understood and used is silly. At best you'd be saving a relatively few people the trouble of looking up this term (you can do that on Wikipedia in about 5 seconds!) whereas it is actually in their interest to learn the meaning of micron if they are going to ever be reading further literature in the area of the Wikipedia article in which it appeared!
Yes, the International Bureau of Weights and Measures (BIPM) did indeed remove "micron" as an official term but more importantly changed the symbol of the micron from μ to the more logical μm which everyone now uses. However in most conversation and writing (99.8% according to my quick poll) "micron" is the term used which is synonymous with micrometer.
Although I don't feel as strongly about it, "centigrade" is also a term for an SI unit whose official name became "Celsius" and needn't be changed for the same reason: it is (essentially) universally understood. Though I will concede that in the last 40 years "Celsius" has come into wide (probably wider) use than its synonym. However for "micron" that absolutely is not the case. It is disingenuous for Lightmouse to state that micron is still used by "some" scientists, carefully adding the disclaimer that "some" means any portion less than 100%. It is the normal term and replacing it with "micrometer" stands out when most people read it. If he (or others) have been going through Wikipedia pages to change the normal terminology to "micrometer," (as he did to one page I had edited) then I believe those unwelcome changes should be reverted. (Is there someone who would know how to program a bot to do that?) Interferometrist (talk) 18:26, 20 October 2010 (UTC)
  • Generally, I like following the BIPM’s rule of the SI. Note however, that Wikipedia goes with the flow and flouts the BIPM’s requirement that a space separate the value and the unit symbol “%”; we write Greg L has a 75% chance of saying the wrong thing on WT:MOSNUM and not Greg L has a 75 % chance as the BIPM would like. In short, Wikipedia does best when it follows the way the world works and doesn’t try to lead by example. I note that “micron” is still commonly used in modern science, notably infrared work in astronomy, as exemplified by this NASA web page. I would recommend a dose of “Follow Current Literature” here. If a clear majority of the RSs in a particular field or subject use “micron” (and I don’t know this to be the case as it applies to infrared astronomy), then Wikipedia should follow that practice. Greg L (talk) 18:37, 20 October 2010 (UTC)
I did a search for micron and micrometer in the IEEE Xplore Digital Library.
  • 1970–2010 : micron = 100,000; micrometer = 29,000 articles
  • 2000–2010 : micron = 55,000; micrometer = 18,500 articles
  • 2010 : micron = 2662; micrometer = 1692 articles
This is a typical one from September 2010.
Yamada, T. (September 2010). "Accurate determination of volume and evaporation rate of micron-size liquid particle". Journal of Applied Physics. 108 (6): pp. 063523–063523-4. doi:10.1063/1.3483250. ISSN 0021-8979. {{cite journal}}: |pages= has extra text (help); Unknown parameter |coauthors= ignored (|author= suggested) (help)
-- SWTPC6800 (talk) 19:57, 20 October 2010 (UTC)
As a general note, micron is disproportionately used in optics and semiconductor devices compared to all other fields of science (much like Ångström is disproportionally used in spectrometry compared to the rest of the science). It's basically a leftover from before the 1960s and the advent of the SI. Prose-wise, writing "a 20 μm switch" is much better than writing "a 20 micron switch" anyway. Now I'm not saying that a bot should make a (micron --> micrometre) switch whenever it encounters it, but micron is obsolete for micrometer. Headbomb {talk / contribs / physics / books} 03:58, 21 October 2010 (UTC)
  • The search above for micron vs micrometer is particularly unfair. The most common term in technical literature is μm. Doing a similar search reveals that μm occurs about 3 times as often as micron. −Woodstone (talk) 05:26, 21 October 2010 (UTC)
@Headbomb: I'm much more used to "a 20 μm switch" than the spelt-out form when reading scientific text. It's as though anyone who doesn't understand the symbol wouldn't understand the full form anyway. And anyone who does understand the symbol would surely prefer it as a symbol. Tony (talk) 08:30, 21 October 2010 (UTC)

On the Corpus of Contemporary American English, even restricting the search to texts from the 2005–2010 period, there are 50 occurrences of micron, 43 of microns, 46 of micrometresmicrometers, and 32 of micrometremicrometer (and a few of the latter refer to the tool). That doesn't sound like obsolete to me. (OTOH, IMO the use of μ alone for the unit is dead: I've only ever seen μm except in publications several decades old.) Conversely, in the corpus and the same period, there are 147 occurrences of Celsius compared to 23 of centigrade, so I wouldn't object to discouraging the latter – providing it's actually used often on Wikipedia, per WP:BEANS. A. di M. (talk) 13:58, 21 October 2010 (UTC)03:06, 22 October 2010 (UTC)

Why did you not include the count of "micrometer", "micrometers" and "μm"? That might change te picture. −Woodstone (talk) 16:42, 21 October 2010 (UTC)
My experience among American semiconductor engineers is that in casual conversation the name of the unit is spoken "micron" but written "μm".
I disagree with the argument above that "micron" is more familiar than "micrometer" or "micrometre"; everyone who understands the basic principle of the metric system, where prefixes are attached to base units, and who has previously encountered "micro-" and "meter", should understand "micrometer". I believe the number of people who would understand "micron" is far fewer.
Greg L's comment about following the literature has merit, but if the literature predominantly uses symbols (μm) as I would expect, how do we reconcile that with the custom in Wikipedia and other non-technical publications to spell out unit names in running text? Should we decide that if "μm" predominates in the literature, it should be adopted by Wikipedia but the symbol expanded to "micrometre" or "micrometer"? Jc3s5h (talk) 17:53, 21 October 2010 (UTC)
I actually searched for the American spellings; that was a typo of mine above. As for the "μm", there are many (most?) people (such as me) who use "μm" as the symbol but "micrometre" for the name. As for Celsius/centigrade and femtometre/fermi, the symbol is completely uninformative as to the choice of full name as there's no other symbol than °C and fm for those units. A. di M. (talk) 03:06, 22 October 2010 (UTC)

Google Search Statistics

So I did a quick search using google to see how many times one or the other term popped up, with the results reported in 1000s of hits. I didn't take into account any overlap between the three terms in the same article (assuming they were usually consistent) but did search using both the singular and plural form. Now, the way Google works, you actually get more hits just searching on "Infrared micrometer" than searching on "Infrared micrometer OR micrometers" which is illogical, probably having to do with what it considers important in your search terms and how many you have specified. But since the three searches in each case used a parallel form, I would conjecture that their results are comparable and the percentage of use of the word "micron(s)" is indicative of the word's actual popularity.

It can be seen that there are some differences in this number between different fields, but that "micron" wins out over "micrometer/re" in each case. Some of the occurances of "micrometer" may have referred to the measuring instrument, and I avoided keywords relating to machining for that reason.

Finally, in every case the use of the symbol μm paired with the keyword was more frequent than any of the searches with the word, but that is not at issue: we all agree on using μm (as mandated by BIPM) rather than the obsolete version μ.

Google search term      Hits × 1000     Percentage using "micron"


Infrared micron OR microns 1380    64%
Infrared micrometer OR micrometers 530
Infrared micrometre OR micrometres 230

interferometer micron OR microns 260    66%
interferometer micrometer OR micrometers 107
interferometer micrometre OR micrometres 28

Grating micron OR microns 1440    90%
Grating micrometer OR micrometers 150
Grating micrometres OR micrometres 11

X ray micron OR microns 1440    65%
X ray micrometer OR micrometers 685
X ray micrometre OR micrometres 99

Cellular micron OR microns 1480    88%
Cellular micrometers OR micrometers 174
Cellular micrometres OR micrometres 23

semiconductor wafer micron OR microns 743    71%
semiconductor wafer micrometer OR micrometers 228
semiconductor wafer micrometre OR micrometres 69

nanotechnology micron OR microns 274    57%
nanotechnology micrometer OR micrometers 141
nanotechnology micrometre OR micrometres 65

protein micron OR microns 1580    61%
protein micrometer OR micrometers 967
protein micrometre OR micrometres 45

No field specified:
micron OR microns 12600    77%
micrometer OR micrometers 3600
micrometre OR micrometres 268
Interferometrist (talk) 18:58, 21 October 2010 (UTC)

  1. Use μm to indicate it, yes.
  2. Link to μm on its first appearance in an article and also on its infobox if applicab;e, as is the current [recommended] practice because the unit is rather obselete.
Also:
  • I've only ever seen it written as micrometre on the unit's first occurence in scientific paper, then μm being used. I've also never heard of it being referred to as "micron" or written just with a μ. Neither have I seen either of these... terms being used in any Wikipedia article. Fortunately.
  • I think British and American spellings have something to do with hits. I'm not very familiar with boolean, but I am sure it has something to do with this.
  • I suppose this is slightly unrelated, but "hard-linking" with just the encoding of the letter in text (μ, not &mu semi-colon) is acceptable. Sorry, the nowiki tag does not work for this parsing stuff.
What's the deal with these posts anyway? It has already been implemented and it is in the Manual Of Style somewhere, I swear I've seen it before. There is no need to patronise people in the talk page, because the talk page is pretty much only for reccomendations on additions and removals to be done to the MOS. And I don't see this information being of any use within the MOS, so can you archive it, already? Thanks, Γιάννης Α. | 20:45, 21 October 2010 (UTC)

In my opinion, micrometre is universally understood to anybody who has a basic knowledge of maths. Micron is far less common, in fact, I've never heard it used except when talking about the width of hairs. McLerristarr | Mclay1 12:01, 24 October 2010 (UTC)

Again, the choice of what units of measure to use in a given article depends on the subject matter. We can’t make project-wide judgements about micrometer or micron based on Google searches on what is found across the entire internet. As I mentioned above, micron appears to be most frequently used by most-reliable, modern sources in the field of infrared astronomy. So it’s simple: For a given topic, adhere to the units of measure used by current, most-reliable sources.
As to the subject of unit symbols like “µm”, our articles, IMO, frequently suffer from “Let’s make it look like a peer-reviewed scientific paper”-itis and we thus find language like …are used for measurements over 2 µm far too soon in an article with insufficient introduction. Few readers of general-interest science articles are comfortable with Greek prefixes like µ. And just because a high-level peer-reviewed journal paper that is referenced in a science-related Wikipedia article directed to a general-interest readership might use “µm”, is certainly no reason for a Wikipedian to do so in order to make our articles as ‘sciency’-looking as possible; Wikipedia articles are not science papers.
Often, our most knowledgable editors use advanced terminology both in articles and on the talk pages because they know their subject matter quite well and are looked up to by others. But, sometimes, these experts in their field aren’t nearly as slick when it comes to technical writing for a general-interest readership; that is, in knowing how best to write in a manner that causes the least confusion and is most accessible to a general-interest readership. Other 16-year-old volunteer wikipedians, seeing a knowledgeable editor use such lingo, will simply follow suit because they don’t want to look like a dolt.
To recap: Note that this NASA site on infrared space satellites is directed to a general-interest readership (not optics scientists) and it uses “micron”. Just because an international, peer-reviewed paper by NASA might be used as a citation in such an article and the science paper exclusively uses “µm”, is no excuse to use it in our articles—particularly when the spelled-out unit name “micron” is so short and doesn’t encumber the flow of reading if it is repeated only a dozen times in the article. For other units of measure with cumbersome, lengthy names when fully spelled out, like mega-electron-volt, properly introducing the unit symbol, MeV, is wise; particularly when the unit symbol is actually pronounced in our minds as “mev”. Greg L (talk) 20:15, 26 October 2010 (UTC)


wp:engvar says "Wikipedia tries to find words that are common to all varieties of English." Similarly, Wikipedia should use terms that are suited to the widest international audience including non-specialists. As Greg suggests, domain specialist terms constantly seeping out from in-house publications into wider domains by 'monkey see, monkey do'. Wikipedia benefits from using SI by default and deviating only if good reasons exist (and hopefully get documented here). Organisations like BIPM (SI units) and NIST (American units) were specifically created to provide stable references for inter-domain, inter-region communication. As Jc3s5h, suggests, anyone that understands 'micron' also understands 'micrometre/er' but the reverse isn't true. I propose the following guidance:

  • Where this guideline is silent on any issue relating to units, use the guidance provided by the BIPM (e.g. it says the symbol for 'kilometre' is 'km' rather than 'Km'). Where the BIPM is silent, use guidance provided by the NIST (e.g. it says the abbreviation for 'foot' is 'ft' rather than 'Ft').
  • Except where required by quotes or text describing history of units:
  • By default, use the term 'micrometre/er' rather than the term 'micron'.
  • In articles about infra-red astronomy, and wool, the term 'micron' may be used if there's some indication on first use that it's a synonym for 'micrometre'.
  • Use the term 'degrees Celsius' rather than the term 'degrees centigrade'.

I hope that guidance encapsulates the issues. Lightmouse (talk) 14:19, 27 October 2010 (UTC)

  • That’s mostly fine with me, Lightmouse. But, really, I think much of it is redundant; MOSNUM already has clear guidance that SI is preferred. Everything we need is succinctly stated at #Scientific and technical units. It includes one, single, all-powerful bullet point that reads as follows:

• Some disciplines use units not approved by the BIPM, or write them differently from BIPM-prescribed format. When a clear majority of the sources relevant to those disciplines use such units, articles should follow this (e.g., using cc in automotive articles and not cm3). Such non-standard units are always linked on first use.

The “micrometer / micron” issue is directly addressed by this guideline, which is a simple principal. Moreover, your fourth bullet point, above, seems like it may be too restrictive; it wouldn’t surprise me in the least if there are other disciplines out there besides infrared astronomy and wool-fiber diameters that use “micron.” There might be many other optic-related articles. If you have a bot trying to make project-wide consistency for spelling, I think it would be better to leave “micrometer / micron” functionality out of it and handle that one manually. That will afford you time to look at the references for that article and gauge what is the best practice. It will also give other editors who specialize in a particular field an opportunity to don their running shoes and catch up with you to discuss things before too many articles have been affected. I think you will find that if a particular article is consistently or usually using “micron” throughout, it is by no accident. Greg L (talk) 18:34, 27 October 2010 (UTC)

You said "MOSNUM already has clear guidance that SI is preferred" and quoted some text from mosnum. Unfortunately it doesn't. If it did, I would have quoted mosnum at Pdfpdf when he swore at me for applying 'km/h' rather than 'kph'. Lightmouse (talk) 18:46, 27 October 2010 (UTC)

I don’t understand why you feel that MOSNUM doesn’t prescribe that the SI is preferred. Note the following, from Which units to use:

*In general, put the units first that are in the most widespread use in the world. Usually, these are International System of Units (SI) units and non-SI units officially accepted for use with the SI; but there are various exceptions for some measurements, such as years for long periods of time or the use of feet in describing the altitude of aircraft.

And from Scientific and technical units:

* In scientific articles, use the units employed in the current scientific literature on that topic. This will usually be SI, but not always; for example, natural units are often used in relativistic and quantum physics, and Hubble's constant should be quoted in its most common unit of (km/s)/Mpc rather than its SI unit of s−1.

Both the above are stating what makes plenty of sense to me and is what Wikipedia is doing: we should generally use the SI unless the normal practice in a given discipline is to use some other unit (or name for a unit). Greg L (talk) 19:36, 27 October 2010 (UTC)
P.S.: As for Pdfpdf (*oops, I just got spittle on my chin pronouncing that*) and his jones over km/h and kph, both are used all over the place. Either side’s quoting world-wide Google hits on one form or the other are of no help. Every article (and its associated topic or discipline) will have its most-common practices. IMHO, the best way to handle it is on a case-by-case basis depending on the most-reliable sources for that article. I know that can be rather frustrating for someone who sees Wikipedia as a project deserving of consistency, but there are a variety of practices out there and Wikipedia does best when it goes with the flow. The art lies in properly identifying what is truly “the flow” for a given discipline; honest wikipedians can have honest differences of opinion when a clear answer is elusive. Greg L (talk) 19:47, 27 October 2010 (UTC)

P.P.S. Speaking of other industries that seem to be still using micron, I note this patent application by Apple for a vapor deposition nitriding coating to stainless steel. It speaks of a coating thickness of “15 microns”. It appears that the use of “microns” is another one of those things on which the the world may be soundly ignoring the BIPM; the proper use of the percent symbol (75% and not 75 %) being another one that immediately comes to mind. Greg L (talk) 19:54, 29 October 2010 (UTC)

Policy on semi-automated edits.

To get this conversation separate from the kph one, the policy is not merely the AWB instructions, it's Wikipedia:Bot_policy. Semi-automated edits that blindly change articles en-masse are covered by the bot policy. Specifically, "Contributors intending to make a large number of assisted edits are advised to first ensure that there is a clear consensus that such edits are desired." If reasonable people are complaining about it, there is not a clear consensus.

"Enforcing" the manual of style with thousands of edits absolutely falls under this. This is why the above conversations counting Google hits for one formatting of a unit vs another are irrelevant. If these changes are controversial enough that they require that sort of conversation, then they should never be done en-masse.

I suggest that complains like pdfpdf's be taken much more seriously. Yes, he degenerated into incivility in the end, but that was after trying to make many good faith efforts to explain that the changes were controversial and unwanted. Gigs (talk) 18:25, 23 October 2010 (UTC)

Ideally this could be achieved (or reversed) simply by adjusting a handful of unit display/conversion templates. ―cobaltcigs 19:49, 23 October 2010 (UTC)
Any possible edit caused by my AWB selections is likely to be opposed by somebody, regardless of the guidelines or the rest of Wikipedia. Sometimes somebody reverts my entire edit, including many unrelated changes, without explanation. But if the bot people want to disapprove my AWB, I guess I'll find something else to do. Art LaPella (talk) 21:04, 23 October 2010 (UTC)
Bots are essential to bringing effect the style guidance to the mass of articles on WP. There is no other way than to bring issues here after someone complains about a run of bot edits. Where they are likely to be controversial (acres to sq m, for example), bot operators need to exercise great caution. But that is no reason to intimidate bot operators in the way that seems to be the theme of the opening comment in this thread. Tony (talk) 02:15, 24 October 2010 (UTC)
Style guidance is just that, guidance. They aren't edicts that should or need to be enforced through mass edits. Often they reflect a very limited consensus of the 4 or 5 people that participated in the MOS talk page discussions. Gigs (talk) 20:44, 24 October 2010 (UTC)
In that case, if style guidance isn't sufficient evidence of a consensus to use AWB, is there any possible AWB edit "that should or need[s] to be enforced through mass edits" based on the AWB user's say-so alone? Or do you want the bot group to micromanage every AWB edit? Art LaPella (talk) 00:50, 25 October 2010 (UTC)
Indeed, there are very few situations where mass semi-automated edits to enforce style are appropriate. It should only be the most uncontroversial of changes. This is our general bot policy, that mass changes need to be very uncontroversial. Gigs (talk) 13:52, 25 October 2010 (UTC)
I don't think that addresses my point. There are 3 kinds of AWB edits, and all of them are likely to be reverted by people who insist they know better: 1. Changes justified by a guideline. 2. Changes with no justification other than the AWB user's say-so. 3. Changes approved by the bot group. So assuming you don't want the bot group to approve everything, why is the first kind more controversial than the second kind? Art LaPella (talk) 00:13, 26 October 2010 (UTC)
Mass changes have a higher bar, whether backed up by a guideline or not. Gigs (talk) 19:58, 26 October 2010 (UTC)
Agreed. Art LaPella (talk) 22:22, 26 October 2010 (UTC)

User:Gigs defines controversy as one query on a talk page. Unfortunately, that definition is retrospective and disproportionate. None of us can predict that an edit will not result in a query from any editor. The correlation between query and controversy is much less than 1:1. Wikipedia would need to ban bots and janitorial editing because no admin or editor can guarantee in advance that queries won't be posted to somebody. I prefer something closer to that at wp:cont: "A controversial issue is one where its related articles are constantly being re-edited in a circular manner, or is otherwise the focus of edit warring. " There are editors that do good work by copy-editing and there are editors that do good work by widespread sub-editing. Consistency and house style actually require many small changes of detail which sometimes makes people with a local article focus ask questions. If sub-editors gave up, nobody would ever get into trouble, but Wikipedia would be worse and certainly less accessible to those without the local article focus. Despite public debates like this one, there's a lot of behind-the-scenes bot, script, and mass editing on Wikipedia and that is a good thing. Lightmouse (talk) 12:27, 24 October 2010 (UTC)

I'm not asking anyone to predict the future. After one person makes a reasonable complaint, you should stop, and take that complaint very seriously. There's a difference between someone asking a question and someone being driven to incivility because a mass editor is ignoring their objections and continuing to make mass edits. Gigs (talk) 20:45, 24 October 2010 (UTC)

Any automated changing of units should be done with care and consensus. Lightmouse has been overriding the default conversion for acres from ha to km2. I think there was a similar issue with nmi. Changes like this are clearly ones that need to be discussed. I will say that finding a good place to discuss these will always be an issue since no matter where it is announced, a good number of editors will be missed. Vegaswikian (talk) 22:43, 26 October 2010 (UTC)

Gigs is correct that bot users should be prepared to correct their many edits if they get them wrong. Lightmouse seems to think that this is unreasonable, as controversial edits cannot be predicted in advance.
Bots allow many edits to be done quickly and easily. The reason why they are supposed to be limited to uncontroversial edits is surely exactly because correcting mistakes is likely to be time-consuming. The risk of having to correct their mistakes is one every bot user takes: like every WP editor they must take full responsibility for their edits. In the case of Lightmouse's edits, several editors have made cogent and forceful arguments against his idiosyncratic views, which have made it clear that the automated edits were controversial. It doesn't matter that this is retrospective: Lightmouse has taken a risk with his idiosyncratic edits, and has come unstuck. He claims that anyone can now go and revert these edits, and that expecting him to correct them is unreasonable. However, this is disingenuous: only a user with the time and technical ability to use a similar bot could keep up with his volume and speed. We now have many automated edits which many editors consider to be wrong: it is for the original editor to revert the edits, or to make them uncontroversial. Not too difficult for someone with Lightmouse's skills and concentration. Richard New Forest (talk) 20:46, 29 October 2010 (UTC)

Distance: 'km' versus 'kms', 'KMs', 'Km' etc

I'd like to do a sub-edit run (unit definition articles and quotes won't be included) to convert synonyms of the symbol of kilometre to lower case 'km', and the symbol of kilogram to lower case 'kg'. . See Wikipedia:Manual_of_Style_(dates_and_numbers)#Unit_symbols:

  • "Symbols have no plural form—an s is never appended (e.g., kg, km, in, lb, not kgs, kms, ins, lbs. Write bit, not bits unless bits is used as a word rather than a symbol)."

I'm being cautious so I'm checking here first and then taking it to BAG. Comments? Lightmouse (talk) 13:09, 24 October 2010 (UTC)

This time I'm being polite. Why don't you leave it for a week or two and let the dust settle?
My impression is that many (including me) are unsettled by your relentless pressure and prodigious speed and volume of edits.
I may have got the wrong impression, but you seem to be largely impervious to similar requests.
Don't get me wrong, we are all impressed by your enthusiasm.
It's just that we can't keep up with you.
BTW: I, for one, very much appreciate that this time you asked BEFORE letting the bot loose.
Best wishes, Pdfpdf (talk) 14:32, 24 October 2010 (UTC)
(P.S. I really do think you should review your priorities. If wikipedia comes out on the top of the list, then I suggest that this is not an ideal situation, and that you may wish to give the "relative priorities" matter further consideration. Pdfpdf (talk) 14:32, 24 October 2010 (UTC))
Pdfpdf, that last comment was completely unnecessary. In answer to the question, there should be no objections to those changes since plural and uppercase symbols have never been acceptable. I'd also like to make the point that even if people sometimes do use unofficial symbols such as "kms" or "kph", that is absolutely no reason to use them in an encyclopaedia. Does AWB automatically correct plural and uppercase symbols? If not, that should definitely be a feature. McLerristarr | Mclay1 15:51, 24 October 2010 (UTC)
"Pdfpdf, that last comment was completely unnecessary." - Dear Mclay1, that's your opinion, and you're entitled to it. I have my own opinion, which is different to yours. Pdfpdf (talk) 10:43, 25 October 2010 (UTC)
Pdfpdf, Lightmouse is also one of the top experts in WP in matters of units and conversions. it is good that he is prompting and participating in debates here. Tony (talk) 16:13, 24 October 2010 (UTC)
Dear Tony1, please advise what I have written that contradicts and/or denies that "Lightmouse is one of the top experts in WP in matters of units and conversions. it is good that he is prompting and participating in debates here." On second thoughts, don't bother. Pdfpdf (talk) 10:43, 25 October 2010 (UTC)
Go for it. I'd suggest there's an argument to be made that pluralisation of some imperial units is pretty common, even officially (British road signs use the symbol "yds", for example, instead of "yd") - but this causes its own difficulties and is not really at issue here. Pfainuk talk 16:28, 24 October 2010 (UTC)
I don't know what a "sub-edit run" is. I don't object to fixing unit symbols that actually are outside of quotations, but I do not believe that bots can reliably detect quotations. Also, since I am familiar with only a tiny fraction of human knowledge expressed in English, I am not prepared to suppose that symbols such as "KMs", "KMS", etc., do not have some proper meaning unrelated to units of measure.
I would sum up my position by asking Lightmouse if he will commit to keeping a log of every such edit, and if any errors are detected in his edits, spend all his time on Wikipedia correcting each and every error he has made, and not do any other Wikipedia activity until all such errors are corrected? Jc3s5h (talk) 16:29, 24 October 2010 (UTC)
Pdfpdf, I concur with other editor's views, that not only is Lightmouse a credible but also highly respected editor, whose contributions in the areas of rationalizing measurements is laudable, consequently, your last "riff" was entirely without merit. FWiW Bzuk (talk) 16:31, 24 October 2010 (UTC)
Jc3s5h, seriously?! FWiW Bzuk (talk) 16:37, 24 October 2010 (UTC).
Considering the low value of minor fixes to symbols, I think it is reasonable for one to commit to clean up any mess one might make. For more important bot edits (for example, complying with copyright laws) a lower standard of accountability might be appropriate. Jc3s5h (talk) 17:46, 24 October 2010 (UTC)
Well no, requiring a separate log of each AWB change would pretty much remove the speed advantage of AWB, which is the main reason to use it, and therefore also remove the most realistic purpose of the guideline. Remember, there are 3 million articles and most editors never see the details of the Manual of Style. Clicking "My contributions" and searching for "AWB" would produce the log you want, assuming most edits include the kind you don't like. My AWB edits already include removing the "s" from "kms" etc., but I don't think that was intended in a sentence like "A bowling ball should weigh 16 lbs." Art LaPella (talk) 19:31, 24 October 2010 (UTC)

As a courtesy for those who care, I'm hereby informing everyone (who cares) that I am busy for the next week and won't be responding to questions/comments/etc in a timely manner. If you REALLY want a response from me before next week, send me email, and then go to my talk page and warn me to look out for your email. (The email account I link to WP gets a lot of junk, which I generally ignore.) Pdfpdf (talk) 10:22, 25 October 2010 (UTC)

  • I’m not sure what this is about, but we should rarely if ever depart from Symbols have no plural form—an s is never appended (e.g., kg, km, in, lb, not kgs, kms, ins, lbs. Write bit, not bits unless bits is used as a word rather than a symbol). I personally disagree with “lbs” since that is extraordinarily common in the the U.S. in grocery-related contexts. But, writing it out (pounds) is infinitely better technical-writing practice so I see no need to make a stink about it. However, certainly for any unit symbol that is part of the SI, one never ever departs from the rule of the SI by adding “s” to denote a plurality. The unit symbol for second is “s”, so “ms” would correctly be the meter·second [millisecond]. Greg L (talk) 20:44, 26 October 2010 (UTC)

This is spell-checking; bots should not be doing it (for precisely the reasons indicated in this section; bots cannot tell when an exception should be made). To quote policy; Bot processes may not fix spelling or grammar mistakes... in an unattended fashion, as accounting for all possible false positives is unfeasible. Lightmouse's record does not encourage me about an attended fix either; but if somebody is willing to check over his shoulder, that would be another question.

But fundamentally, why do this with a bot at all? If there is project-wide consensus for this fix, the normal process of fixing errors as they are noticed will take care of it, as it takes care of teh; if not, it should not be done at all. Septentrionalis PMAnderson 17:12, 27 October 2010 (UTC)

  • We were poking our fingers around in a little whirlpool that had caught a tadpole of an issue in its currents (an issue of not using plural ‘s’ for unit symbols) and you, PMA, just did one of your cannonballs into the water just two inches away over Lightmouse’s bots. (*shaking water out of my hair*) Lightmouse’s bots (Lightbots), as far as I know, are all pre-approved over in botland—wherever that is; I don’t remember the name at the moment. Please take your issue there. Greg L (talk) 18:13, 27 October 2010 (UTC)
When you find out where the botland approvals are, do let us know; the proper answer to the question at the top of the section remains negative, here or there. Septentrionalis PMAnderson 20:39, 27 October 2010 (UTC)
You’ll have to ask Lightmouse. I recall seeing some posts where he was seeking approval for one of his bots. One editor over there was promising regular rectal exams with an 8-centimeter speculum. Greg L (talk) 21:29, 27 October 2010 (UTC)
Wikipedia:Bots/Requests for approval. And of course the project-wide consensus for things like "kms" is "huh?" Apparently a tiny minority wants "km", but the minority wanting "kms" is much tinier, so this sounds like a good time for a bot if the false positives can be minimized. Concerning "lbs", perhaps it should ideally be "pounds", but Wikipedia gets 14,500 hits for "lbs" (minus a couple thousand irrelevant hits). So we definitely shouldn't be changing "lbs" to "lb" unless we want it in places like Stan Efferding: "Efferding started competing in powerlifting in 1996 totaling over 2,000 lbs in the Pepsi Region 8 challenge". Without the "s" I would mentally read that as "totaling over 2,000 pound". Art LaPella (talk) 23:18, 27 October 2010 (UTC)
To clarify myself: We have too many rules, but forbidding automation is more like adding a rule than removing one. Art LaPella (talk) 04:08, 28 October 2010 (UTC)
  • Someone please tell me that “kms” is being fixed by a bot. Appending a plural-“s” to a unit symbol that is part of the SI is just soooo wrong. Whomever constitutes this minority voice advocating such a practice simply has no idea about proper technical writing practices. Greg L (talk) 22:35, 29 October 2010 (UTC)

Celsius

It appears uncontroversial that we should assist editors by adding the final sentence below.


Your thoughts, please? Lightmouse (talk) 13:52, 28 October 2010 (UTC)

A Google search for "degrees centigrade" on en.wikipedia.org only shows 128 results, some of which are probably quotes, a discussion of historical units, or not article pages. This problem seems quite small compared to other examples of poor writing, so I don't think it merits mention in this manual of style. Jc3s5h (talk) 14:00, 28 October 2010 (UTC)
But the "coulomb" and the "farad" are worth mentioning? This additional sentence seems reasonable to me. Tony (talk) 14:17, 28 October 2010 (UTC)

It's one of those colloquial words that people add from time to time. Trust me, there are editors out there that sub-edit articles to replace such terms and it'd be worth having a guideline to make it clear in case somebody ends up swearing at them. Lightmouse (talk) 14:33, 28 October 2010 (UTC)

Seems like unnecessary bloat to me. Is this actually a problem, or are we talking of hypothetical scenarios? Because I very strongly doubt that anyone would get there panties in a bunch if someone switch centigrade to Celsius (where appropriate), and whoever objected couldn't possibly gain consensus for using centigrades. Headbomb {talk / contribs / physics / books} 18:59, 28 October 2010 (UTC)
"Degrees centigrade" is evidently hypothetical. Since there are only 128 results (130 when I looked), I checked them all. There are talk pages, Wikipedia namespace pages, Wikiquote and Wikibooks, quotes, and cases where "degrees centigrade" has been reworded even though it still shows in Google. But I didn't find one single example that should be changed. Someone else must already be searching Wikipedia for that phrase. Art LaPella (talk) 21:59, 28 October 2010 (UTC)

There's only 128 because editors like me keep changing it to Celsius. I do a run from time to time and even in my routine work, I pick them up. There are other editors that do it too. Google shows centigrade in 7 to 25% of hits (depending how you search), so it's not surprising that it finds its way into WP articles. I found the following instances just now:

We have some astonishing trivia in explicit detail on mosnum, yet routine stuff isn't. Gnoming editors are in a Catch 22 where one person can swear at them because mosnum doesn't document consensus for <insert standard term or format> but if the consensus is strong, it won't be documented. The way things are right now, we need to vote on everything. Lightmouse (talk) 11:41, 29 October 2010 (UTC)

  • I'm a bit of an armchair style-guidist, and I think what LM does is impressively wider than that: he actually exposes himself to the real world of the frustratingly complex, sometimes messy, undecided and inconsistent usage of units on WP. For this, he cops flack, and is in an ideal position to let us know if he and others like him who do the hack work of cleaning up this aspect of the project need the support of bits of text in MOSNUM here and there. I think we should support him in this respect; at the same time, I'm not suggesting that we should not scrutinise his requests and his work. Tony (talk) 12:31, 29 October 2010 (UTC)
What he said. Bzuk (talk) 13:30, 29 October 2010 (UTC)
Perhaps some of the mind-numbing trivia in the MOSNUM, which serves primarily to back up editors who make corrections, should be removed to a subpage, maybe Wikipedia: Manual of Style (decisions on dates and numbers). Jc3s5h (talk) 14:21, 29 October 2010 (UTC)
Jc3, I've no in-principle objection to the segregation of more arcane or specialist material into appendices. The page might work better then as a way for editors to learn about units, etc., me included. But until that happens, when there is evidence by someone at the coal face that there's a continual trickle of additions of an undesirable and non-SI unit, why can't this be covered briefly in the guideline? Tony (talk) 15:28, 29 October 2010 (UTC)

(arriving late to this discussion) Yes. “Centigrade” is beyond archaic and should only be used in direct quotes or when talking about the history of the temperature scale. Greg L (talk) 19:59, 29 October 2010 (UTC)