Template talk:Weather box/Archive 3

Shouldn't the template default to metric?
It would appear that the template defaults to Imperial unless the line: |metric_first=yes is placed in it.

Given that metric usage is much more widespread than Imperial usage, shouldn't it be the other way around and default to metric first, rather than Imperial? Nfitz (talk) 21:32, 25 March 2009 (UTC)


 * I would support that. --Joowwww (talk) 21:50, 25 March 2009 (UTC)

Metric is not more widespread than Imperial in English-speaking countries, and this is the English Wikipedia.--Paul (talk) 21:56, 25 March 2009 (UTC)


 * WP:MOS gives the system of measurement to use as that which corresponds to the article's subject. So according to WP:MOS articles about any country other than the United States, Burma and Liberia would have metric measurements. I can only assume that there are more articles about metric countries than imperial countries. --Joowwww (talk) 22:17, 25 March 2009 (UTC)


 * Metric isn't more widespread than Imperial?? Good grief, I've spent a lot of time in 7 different English-speaking countries, and metric was dominant in all of them except one - especially for outdoor temperatures!  I've spent all my life in English-speaking countries, and I can barely understand the Farenheight measurements - I can convert, but the temperatures 50, 60, 70, 80 mean nothing to me until I convert it to English. Nfitz (talk) 01:20, 26 March 2009 (UTC)


 * I think he means that there are more English speakers in the US than the other English-speaking countries. Although when you take into account the world's English speakers, whether native or not, which the English language article gives at a maximum of 1.8 billion, imperial users are certainly in the minority. --Joowwww (talk) 12:55, 26 March 2009 (UTC)


 * English language indicates that there are not more English speakers in the US than other countries. Besides the language is called English not Yank Nfitz (talk) 04:09, 6 April 2009 (UTC)

(outdent) More than one thousand pages use this template with its current default. Who is volunteering to fix all of those pages which will break if the default is changed?--Paul (talk) 04:16, 6 April 2009 (UTC)


 * It would be a case of adding  to those that are used on US pages. Those that aren't US templates wouldn't be affected as the   parameter would become redundant. People can add the parameter to articles they edit when they notice it's in metric, it can be done in no time. I'm pretty sure Wikipedia has seen bigger changes. --Joowwww (talk) 10:45, 6 April 2009 (UTC)


 * If that's all it is a bot could do it. It could even be done with AWB. Enter CambridgeBayWeather, waits for audience applause, not a sausage 13:35, 6 April 2009 (UTC)


 * If this is so, then what is preventing us now from changing the default to suit the majority? -- Hamster   X  06:01, 9 May 2009 (UTC)


 * There's no reason to use the non standard imperial unit as a default. As pointed out above WP:MOS states "International scope: Wikipedia is not country-specific; apart from US and some UK specific topics, use metric units." Please add  and do the   parameter redundant. Nsaa (talk) 08:44, 15 July 2009 (UTC)


 * Is there any movement on changing the template to default to metric? Metric measurements predominate for weather reports everywhere except the US, and since WP:UNITS requires a general preference for metric anyway, it's a no-brainer. Mixsynth (talk) 19:11, 23 December 2009 (UTC)


 * Since the general consensus here appears to be that the unit default of this template should be metric, I intend to change it myself so that the default is metric and the toggle parameter is "|imperial_first=" rather than "|metric_first=". Before I do so, I would appreciate feedback from others here as to whether or not a bot should first be employed to edit all existing "imperial first" implementations to include the new "|imperial_first=yes" toggle parameter, or whether it should simply be left for individual authors to notice the change and add the "|imperial_first=yes" parameter themselves if they wish. Mixsynth (talk) 17:18, 10 January 2010 (UTC)


 * I agree that it should default to metric, but I think the parameter name should be something like  with options   and  . Then the template won't be as vulnerable to changes in the defaults. Thomas Nygreen (talk) 16:12, 16 January 2010 (UTC)

Name and purpose of fields
"MMM_Hi_°F" and "MMM_Lo_°F" are, in the resulting table, labelled as "Average high" and "Average low", but it's not clear to me what this means. Does it mean the average of the highest/lowest temperature recorded over the whole month? Or does it mean the average of the daily highs/lows over the given month? I think this needs clarifying in the documentation and in the labelling. 86.134.10.88 (talk) 21:51, 6 January 2010 (UTC).

False precision in snow conversion
It appears this template converts 7.5 inches of snow to 190.5 mm of snow. 7.5 inches of snow is precise to a tenth of an inch. 190.5 mm is precise to a tenth of mm - which is 1/254th of an inch -- over 25 times more "precise" than the data being converted. This is false precision & snow conversion should default to nearest mm or nearest tenth of a cm.--JimWae (talk) 06:16, 21 March 2010 (UTC)
 * Assuming that the inches is obtained in the equivalent way that other countries obtain their snowfall, and a look at Juneau, Alaska would seem to confirm that, it should be converting to cm. something lame from CBW 13:26, 21 March 2010 (UTC)

In Canada, where rainfall is measured in mm, snowfall is measured to the nearest cm. It is only averaging that results in tenths of a cm for snowfall. The false precision problem exists for both NYC & Juneau weather charts (and probably many others), neither of which has any raw "input" data in "metric". Thus the template must be converting by default with false precision. I cannot see where this template makes the conversion. Can anyone figure out where to fix this?--JimWae (talk) 20:11, 22 March 2010 (UTC)
 * Using "tenths" in the averaging is false precision, too. It might not be introduced by the template, but it still isn't proper. Gene Nygaard (talk) 14:16, 23 March 2010 (UTC)
 * That's not correct. Snowfall in Canada is measured to the nearest 0.2 cm and rainfall to the nearest 0.2 mm and neither are averaged. The averaging comes in here where multiple years are used to get the normals. In the US snowfall is to the nearest tenth of an inch and rainfall to the nearest one-hundredth of an inch. The template should be converting to tenths of a centimetre for snow and tenths of a millimetre for rain. something lame from CBW 16:06, 23 March 2010 (UTC)
 * I think that should fix it. 117Avenue (talk) 06:30, 24 March 2010 (UTC)
 * However what you have now is not much better. Snow is measured to the 1/10 of a centimetre or 1/10 of an inch. Rain to 1/10 millimetre or to 1/100 of an inch. What we now have is both rain and snow being converted from US measurements to the nearest cm and mm. But it's converting metric to 1/100 of an inch (rain) and 1/10 of an inch (snow). something lame from CBW 05:59, 25 March 2010 (UTC)
 * This is odd. Look at Climate of Miami the precipitation is converted to 1/10 mm. But Climate of Dallas is being rounded to the nearest mm. That's OK for January 1.89 in but February 2.31 in. something lame from CBW 06:14, 25 March 2010 (UTC)
 * Miami has input data in "metric" -- as tenth of mm--JimWae (talk) 09:10, 25 March 2010 (UTC)
 * Miami is input tenth of a mm, so there is no false precision there. Dallas is input hundreth of an inch, a hundreth of an inch is between a tenth of a mm and one mm, so converting to a tenth of a mm would be false. 117Avenue (talk) 14:56, 25 March 2010 (UTC)
 * Several issues here. One, at the low end of the scale, where precipitation is less than 10 mm (0.39 in), and input is imperial, the precision of metric conversion needs to be to 0.1 mm (look at Phoenix. Second, the metric conversion for snow needs to be to the nearest mm (0.1 cm); there is less of an issue, as 0.100 in = 0.254 cm. 2.54 times the precision is much less problematic than 25.4 times, as was the case previously, where snow conversion was to nearest 0.1 mm. 华钢琴49 (TALK) 15:11, 25 March 2010 (UTC)
 * Converting hundredth of an inch to tenth of a mm is false! For example, the precipitation in Phoenix in June is 0.09 inches, is that 2.2mm or 2.4mm? If you want Dallas and Phoenix to be like Miami, then use metric. 117Avenue (talk) 23:47, 25 March 2010 (UTC)


 * US boxes should not be in metric and most certainly not in a mixture of the two. At least it's sourced. However, that means that the WMO is converting 1/100 of an inch to 1/10 of a mm. something lame from CBW 00:39, 26 March 2010 (UTC)


 * good point, but with regards to snow, the colouring scheme for precip of any kind (total precip, rain, or snow) is determined by 10 mm increments. The purpose of this template is to provide, via colouring, a visual sense of how heavy precip is or how high temperatures are. Rounding to the nearest cm with snowfall somewhat defeats that purpose. 华钢琴49 (TALK) 22:04, 26 March 2010 (UTC)
 * The colouring scheme still works, but do you propose a new scheme for snow? It maxes out for Canadian cities. 117Avenue (talk) 23:46, 26 March 2010 (UTC)
 * If, in precipitation, there is only 1 decimal point inputted (which I hope is rare), then by your standard, the mm conversion should be to the nearest 10 mm, as in that scenario, rounding to the nearest mm 'creates' 2.54 times the precision.

Besides, 'creating' 2.54 times the precision is much better than 25.4, better than even 10 times. And at the severely low end of the scale for precip (a few mm), .1 mm is far more significant. Regarding snow, I did not say that rounding to the nearest cm causes the scheme to falter. I said that rounding to the nearest 1 cm, when the colouring is based on 1 cm increments, makes the scheme pointless. and yes, there should be a new scheme, as 15 cm (6 in) of snow in a month is not much for snow regions; think about it... all you need is a few light snowfalls in a month. 华钢琴49 (TALK) 20:17, 27 March 2010 (UTC)


 * example of what I am complaining with regards to snow. in the Washington, DC article: BOTH 1.6 in and 1.4 in map to 40 mm? mathematical/statistical blasphemy! 华钢琴49 (TALK) 05:12, 29 March 2010 (UTC)
 * I did not mean to start an edit war. The chages you made broke it, and induced an error. I was simply reverting to a format that worked while I wrote and debugged the changes you have proposed. You changed my mind, the inch precision will now make the same cm precision. 117Avenue (talk) 08:03, 29 March 2010 (UTC)
 * Ok. I will leave it to you and others then. I did not realise that I induced an error, and am sorry for that. 华钢琴49 (TALK) 14:15, 29 March 2010 (UTC)

cm vs mm
Rain fall is usually given in mm rather than in cm. Snowfall is usually given in cm. At least that is the case in Canada. Peter Horn User talk 16:00, 23 March 2010 (UTC)
 * See above in both cases it's measured to the nearest tenth. something lame from CBW 16:08, 23 March 2010 (UTC)
 * Thanks. I was referring to Crete. Peter Horn User talk 16:24, 23 March 2010 (UTC)


 * This may be true in Canada. Because of the large number it produces outside of desert areas, I would think that centimeters would make much more sense. cm is "comparable" to inches just as "yards (or feet)" is "comparable to "meters." If the use of mm is to discourage Americans from switching to the metric system, by all means leave it! :)  Student7 (talk) 12:43, 18 May 2010 (UTC)


 * I'm not sure what you mean but mm is the standard for most of the world. See Australia, France and the UK for some examples. Enter CBW, waits for audience applause, not a sausage. 15:44, 18 May 2010 (UTC)

Slight problem
It appears that the template is not usable unless you have average high and low for each month. It should work with only the mean temperature or without temperature at all. something lame from CBW 22:04, 7 April 2010 (UTC)


 * Who made the decision that that average high and low were mandatory? Is it required to determine if the year column is needed? 117Avenue (talk) 23:12, 7 April 2010 (UTC)
 * I have no idea. But it should work with the use of just one group and no group should be mandatory. On the other hand it can't be that much of a problem if it only came up now. something lame from CBW 07:06, 8 April 2010 (UTC)

✅, fixed. 117Avenue (talk) 05:28, 11 June 2010 (UTC)

Temperature
I suggest new colors for the temperatures:

I do not agree with the current system. Eduardo Sellan III (talk) April 23, 2010, at 09:05 pm.


 * Why not? The extreme temperatures of earth are -89°C and 58°C, the colors need to be extended to those temperatures. 117Avenue (talk) 00:16, 24 April 2010 (UTC)


 * I disagree because I believe that the colors do not represent the temperatures as they should. Eg, 10°C is represented in a low tone of orange, which makes the temperature seem hot. But in most cities of the world it is considered a low temperature. The extremes outside the standard of -59 to 60 would be black. Eduardo Sellan III (talk) April 23, 2010, at 09:25 pm.


 * That would make all the temperatures between -89 and -59 the same color, which I disagree with. There are cities where 10°C is warm, orange doesn't mean hot, 20°C is still redder than 10°C. How about this:<!--


 * 117Avenue (talk) 04:30, 26 April 2010 (UTC)


 * I agree. I thought a little, and concludes that its proposal would represent very well the differences between cold and heat. Eduardo Sellan III (talk) April 27, 2010, at 11:05 pm.

But I think you should improve the picture, putting the colors two degrees later. For example, the color of 6ºC should represent 8°C. Eduardo Sellan III (talk) April 30, 2010, at 7:57 pm.
 * I will perform the change in a couple of days, I am working on other changes right now. 117Avenue (talk) 23:24, 30 April 2010 (UTC)
 * You are suggesting that 6.5°C be represented by white rather than 4.5°C. I don't think it should be any farther from 0°C, because white means ice and snow, which can only exist below 0. I think that 8°C is a good shade, redder than 5, and considerably cooler than 20. 117Avenue (talk) 02:56, 2 May 2010 (UTC)
 * I agree, snow does not occur after 7ºC, and this is the last temperature represented in white. Eduardo Sellan III (talk) May 2, 2010, at 6:14 pm.


 * Snow can fall and exist on the ground even if the temperature is above freezing as well as melting when the temperate is below freezing. The highly esteemed CBW presents the Talk Page! 00:48, 3 May 2010 (UTC)
 * So what are you saying, where do you think white should be? 117Avenue (talk) 01:24, 3 May 2010 (UTC)


 * Well if it was up to me I would remove all the colours and go with black on white. That way there are no problems what people are using to view Wikipedia. However, I realise that is not really an option. The thing is that depending on which monitor, of 5, I'm using, 2 at home different makes but off the same video card and 3 different ones at work the colours all look different. Right now, It appears that white is used for +1 - +5 but as I look at the chart the colour from 0 and +6 bleeds in. So after about 30 seconds white is only visible on +3 and +4. Is the same blue used through -27 to -29? For some reason the black on blue for -28 and -29 is hard to read. The highly esteemed CBW presents the Talk Page! 21:12, 3 May 2010 (UTC)
 * My new system is formula based, so that every colour is different, the bleeding was my purpose for this change. -29ºC is #4A4AFF, -28ºC is #4F4FFF, -27ºC is #5555FF, 4.4ºC is #FEFEFF, 4.5ºC is #FFFFFF, and 4.6ºC is #FFFFFE. The problem I was having figuring out at what value the font colour should change, was that over a range both options don't look good, and it will depend on the monitor the user is using. Perhaps it should be a lighter blue. 117Avenue (talk) 00:51, 4 May 2010 (UTC)
 * I'll fix the font colour problem, I was using a bright laptop screen to write it, but I see what you mean when I look at it on a monitor. 117Avenue (talk) 05:55, 13 May 2010 (UTC)

✅ 117Avenue (talk) 05:28, 11 June 2010 (UTC)

Although frost is possible at above freezing temperatures, light blue colouring shouldn't even begin until -0.1 °C. ---华钢琴49 (TALK) 18:36, 12 June 2010 (UTC)
 * As CBW said above, it can snow when its warmer than 0°C, and melt when its less than 0°C, the figure of 4.5°C for white stood unopposed for a month, but I can try to write a scheme with blue and red blending into each other, removing the white section that CBW also noted. 117Avenue (talk) 04:15, 13 June 2010 (UTC)

Precipitation
Currently the precipitation colouring scheme (colp), colours everything past 150mm of rain, and 15.0cm of snow the same colour. I propose that it be extended to 200mm of rain like this: My question is, should snow use the same colours as rain and precipitation, or should it be ten times smaller (i.e. 90cm of snow = 90mm of rain)? Snow only equates to ten percent of rain, and applying a factor of 0.1 to the snow colour, would make the precipitation colour appear to be the sum of the rain and snow colours. For example, currently for Edmonton the snowfall in January is a deep blue, and the rainfall is near white, but the sum is a light blue, this appears to be an addition mistake. 117Avenue (talk) 05:20, 3 May 2010 (UTC)


 * because of standards of what defines 'large' snowfall averages, snow should definitely not use the same colours as rain/precipitation. for example, the 5.9 in of snowfall in January in Washington, DC shows as 15 cm and thus uses the darkest colour, but 15 cm of snow in one month is not much. However, applying the '0.1 factor' tends to the other extreme; 90 cm of monthly snowfall should be considered heavy, and most snow-receiving places in the world cannot at all think about receiving that amount in one month (yes, for a season, but...). I don't know where to set the minimum snowfall value for the darkest value. the shading intervals should depend solely on that value. --- 华钢琴49 (TALK) 12:59, 3 May 2010 (UTC)


 * For the sake of argument assume that 1 mm uses the colour white. So both rain and precipitation would use white for 1 mm. Then due to the 10:1 conversion for snow to rain 10 cm of snow would also be white. The highly esteemed CBW presents the Talk Page! 21:32, 3 May 2010 (UTC)
 * By the way can you change the 70 and 80 to be white on blue. The black is hard to read. The highly esteemed CBW presents the Talk Page! 21:34, 3 May 2010 (UTC)

The question over daily rate has been brought up, but only by using the calculation monthly precip / number of days in month. why not, then, calculate the daily rate by taking the monthly precip over the number of precip days and altering the colouring based on that? 华钢琴49 (TALK) 04:01, 4 May 2010 (UTC) Extending it to 300mm, and using it for snowfall as well. As I look at various cities, the max rain and max snow seems to taper off after 250mm for both. 117Avenue (talk) 06:51, 4 May 2010 (UTC)
 * I don't see how that would be useful information. How is this second option:


 * Sorry for not being clear once more.


 * 'I don't see how that could be useful information': Again, look within this talk page for discussion on daily rates. If 196 mm of precip falls in January, and 184 mm in February, then which is the wetter month? Obviously February, since it's daily rate is 6.51 mm, as compared to January, with 6.32 mm/day.
 * could we change the colours to green or have a parameter allowing users to switch to green?
 * extending it to 300 mm is reasonable. I would have preferred 250 mm, since that is just short of 10 in, but since you have already coded that colouring scheme, whatever. but if you wish/have the time, you could adjust the limit to 250 ---华钢琴49 (TALK) 21:25, 17 May 2010 (UTC)


 * I wouldn't reduce the 300 mm to 250 mm. While I'm not sure how reliable this or this it certainly indicates that while the monthly amounts, Astoria, Oregon and Mount Washington (New Hampshire), are fine the yearly amounts may be the same colour as the monthly amounts if a smaller scale is used. Then of course there's Andagoya, Colombia, Andagoya. Enter CBW, waits for audience applause, not a sausage. 22:44, 17 May 2010 (UTC)
 * Okay, I understand now, I can write in a multiplication factor for colours for months not 30 days long. Did you mean you want green to represent precipitation? 117Avenue (talk) 23:44, 17 May 2010 (UTC)


 * that would be nice, but first wait for input, as changing the colour is a major change. But keep the snow colours. ---华钢琴49 (TALK) 17:43, 18 May 2010 (UTC)
 * ✅, scprecip and scrain can now be used to make green colours. 117Avenue (talk) 05:28, 11 June 2010 (UTC)

Humidex and windchill
Currently the humidex and windchill values use the humidity colouring scheme (colh), which is currently white for all values. I propose that they both use the temperature scheme, humidex be moved above record high, and windchill be moved below record low. 117Avenue (talk) 05:20, 3 May 2010 (UTC)
 * ✅ 117Avenue (talk) 05:28, 11 June 2010 (UTC)

Humidity
Currently the humidity colouring scheme (colh) is white for all values. I propose this scheme: I think that it will give the impression of a wetter atmosphere with the higher values. 117Avenue (talk) 05:20, 3 May 2010 (UTC)
 * ✅ 117Avenue (talk) 05:28, 11 June 2010 (UTC)

Sunshine hours
117Avenue, could you revert your changes to the sunshine colouring scheme? The annual sunshine average for Lhasa now shows as a dull yellow colour, instead of the bright colour it used to be; 2990 hours of sunshine annually is a LOT, and the new colour suggests that it isn't. ---华钢琴49 (TALK) 13:36, 3 May 2010 (UTC)

also the Chicago infobox is now garbled up; apparently the sunshine parameter will not accept percentage values. --- 华钢琴49 (TALK) 21:34, 3 May 2010 (UTC)


 * I'd noticed the same thing where n/a or was used. I left a message at User talk:117Avenue as well. The highly esteemed CBW presents the Talk Page! 22:10, 3 May 2010 (UTC)
 * Why the heck are there percentages being used in Chicago?!? Only numbers are to be entered, it was erring before, because the text should to be 80%, not 100%. The current colouring scheme colours everything over 2,730 hours of sunshine a year yellow, however 12 hours a day for 365 days a year is 4,380, I propose that the colouring scheme be extended like this:
 * So that yellow is 4,372, and white is 8,743. 117Avenue (talk) 00:51, 4 May 2010 (UTC)

It should be closer to what it was previously. 117Avenue (talk) 06:51, 4 May 2010 (UTC)
 * oppose. It should stop at around 4,380 hours, which is the calculation for when South Pole receives 100% sunshine, which is not very likely, in its 'summer months'. 2500 hours annually is a lot for many, many places around the world, and should thus be represented by quite a bright colour, and 3000 even brighter. I suggest moving the colour you proposed for 4000 to 2500, and 4500 to 3000. --- 华钢琴49 (TALK) 02:38, 4 May 2010 (UTC)
 * I have inserted the #iferror function, this will prevent the error in all the articles that have text for values. This scheme moves olive (#808000) from 2,186 to 1,093:


 * Thanks for fixing that. The highly esteemed CBW presents the Talk Page! 08:20, 4 May 2010 (UTC)

Why not base the sunshine colouring scheme for each month on the mean daily sunshine duration? ---华钢琴49 (TALK) 15:03, 9 May 2010 (UTC)
 * I'm not sure what you're saying. The colour scheme will be applied to other scales, yellow is assigned to 364.4 hours per month, which is equal to 4,372 hours per year, or 11.98 hours per day. 117Avenue (talk) 05:55, 13 May 2010 (UTC)
 * most prominent difference is between January and February. Say January receives 5.4 hours of sunshine per day, and February 5.9 hr; thus the monthly total would be 167.4 hr for January, and 166.7 hr for February. Though, based on monthly sunshine totals, not daily averages, it may appear that January is sunnier, February clearly is. ---华钢琴49 (TALK) 15:23, 13 May 2010 (UTC)
 * Climate data is available in hours of sunshine per month. Sure, we could have a row for average number of hours per day, for each month, but that data isn't available, at least not in Canada. 117Avenue (talk)


 * sorry for the confusion. either I am not being clear enough, or you do not understand. I am suggesting that the colouring be changed based on the daily sunshine amount, not the monthly total. this option has been discussed with precipitation before, in the section 'Misleading colours with precipitation' or whatever it is named. ---华钢琴49 (TALK) 21:10, 17 May 2010 (UTC)
 * Okay, I understand now, I can write in a multiplication factor for colours for months not 30 days long. 117Avenue (talk) 23:44, 17 May 2010 (UTC)


 * 117Avenue, I looked at the code for COLS, and copied all of the beginning portion (not the table) onto the Chinese wiki to update the colouring, but the annual sunshine amount is not shaded correctly. So do you know where the code for the annual amount is? ---华钢琴49 (TALK) 15:40, 3 June 2010 (UTC)
 * Sorry I haven't been able to implement the changes yet, it has taken way longer to write then I thought. It is Template:Infobox weather/line/onevalue that divides the year value by 12 (12.175 after future code change) in the COL template. 117Avenue (talk) 01:13, 4 June 2010 (UTC)

✅ 117Avenue (talk) 05:28, 11 June 2010 (UTC)

3rd coding option redundant
Why not delete the 3rd (metric and imperial input required) coding option? it is far more time-wasting than coding the other two, because it requires each user to manually convert the figures. Users can still use 'mixed' coding, where some data is inputted using metric, and some imperial. 华钢琴49 (TALK) 21:02, 1 May 2010 (UTC)
 * Sorry, I don't understand what the problem is. What values have to be manually converted? Metric or imperial can be inputed, even in the same row, and it will always appear in the same order in the cell. 117Avenue (talk) 02:56, 2 May 2010 (UTC)


 * I am referring to 'imperial and metric empty syntax', the last option of the template, which states 'when one parameter of one system is filled in then the corresponding parameter of the other system must be filled in'. unless this sentence is completely false, the problem is that far more work is required on the part of the user. of course the other issue is that it creates additional space here in the template. 华钢琴49 (TALK) 04:18, 2 May 2010 (UTC)
 * Thanks for pointing that out, it is a mistake. The third option is included in the documentation for those who chose to put in metric for some things and imperial for others, or to see the full syntax. 117Avenue (talk) 05:58, 2 May 2010 (UTC)


 * There's another section on this above at Template talk:Infobox weather. The highly esteemed CBW presents the Talk Page! 21:30, 3 May 2010 (UTC)

Precipitation days colouring
should obviously stop at 31 days. there aren't 31.6 days per month. I am only alerting this, as I am a n00b wrt colouring ---华钢琴49 (TALK) 03:23, 24 May 2010 (UTC)
 * I know, I just wanted the table to be ten cells wide, and 31.6 was the closest I could get to 31. This will be changed though when the length of month factor is applied, 31 days in a 31 day month, and 28 days in February, will be the same colour as 30 days in a 30 day month. 117Avenue (talk) 04:08, 24 May 2010 (UTC)

Second Request: 117Avenue, could you switch the coding of precipitation day colouring so the hexadecimal template does not have to be used? I am trying to allow Chinese Wikipedia access to precip, rain, snow days, but my Chinese is not high-level enough to translate all of the hexadecimal template documentation; I am initially using the old colouring scheme, where the colouring was based on increments of 5 days, which is dumb. ---华钢琴49 (TALK) 03:50, 24 May 2010 (UTC)
 * The purpose of using the hexadecimal template, and equations, was so that every number would be different, and the coding would be shorter. If you don't want to write the necessary templates on the Chinese Wikipedia, you'll have to calculate the colour code for every number. 117Avenue (talk) 04:08, 24 May 2010 (UTC)


 * ok. both responses are reasonable. for the second request, was just asking you if there was another existing method to code in the colouring scheme. Still thanks for the work on the gradual colouring though. ---华钢琴49 (TALK) 20:59, 24 May 2010 (UTC)

Minus signs
Is it possible for the template to use minus signs instead of hyphens for negative temperatures? Dabomb87 (talk) 00:25, 27 May 2010 (UTC)
 * Sure, you can type a minus or a hyphen into the template on the article. In fact minus is encouraged by MOS:DASH. 117Avenue (talk) 00:42, 27 May 2010 (UTC)
 * Yes, but some of the conversions' dashes cannot be controlled manually. For example, see Template:Grand Forks weatherbox (which uses this template as its base); all of the negative Fahrenheit temps use minus signs, but their corresponding Celsius conversions do not. Dabomb87 (talk) 02:11, 27 May 2010 (UTC)
 * That would require the parser function to render &amp;minus; instead of a hyphen, which it probably should do anyway since it accepts the minus sign as an input argument. In any case, that's beyond the scope of this (or any) template; it's part of the MediaWiki syntax. — Andrwsc (talk · contribs) 04:18, 27 May 2010 (UTC)
 * That's right, conversions are done by #expr. If you'd like, you could convert them yourself, and input both celsius and fahrenheit. 117Avenue (talk) 04:42, 27 May 2010 (UTC)
 * Kinda defeats the purpose of automatic conversion by this template.... — Andrwsc (talk · contribs) 21:35, 27 May 2010 (UTC)

Sea temperature
In infobox there is one option lacking - Sea temperature. Increasingly, such information shall be provided by Meteorological Agencies. Subtropical-man (talk) 19:02, 30 May 2010 (UTC)
 * obv not widely available today, but due to supposed 'global warming'. Let's not add it until an agency provides it for every maritime location. ---华钢琴49 (TALK) 15:36, 3 June 2010 (UTC)
 * I don't think that sea temperature is really weather related. 117Avenue (talk) 01:13, 4 June 2010 (UTC)
 * This option works well on other Wikipedias, example in Russian (Температура воды): . In addition, increasingly such information shall be provided by Meteorological Agencies. Time to make this option here, on English Wikipedia. Subtropical-man (talk) 13:06, 8 June 2011 (UTC)

Colouring scheme
117Avenue, regarding your colouring changes:
 * Temperatures: I approve of the degree-by-degree colouring scheme, but our standards of what is cold and what is hot have been set too high. It appears that the current colouring for 30 C is what approximately what was used for 24-27 C. Make no mistake, 30 C is HOT for much of the world, and it, not 38 C, should be the beginning of the red colours.
 * Liquid (Precip, humidity, precip days): very well done! I especially appreciate the addition of the green option. If, for precipitation and rainfall, you could allow the user to, in one fell-swoop (not twice), apply green colouring, that would be excellent.
 * Sunshine: Please revert these edits. A while back, you agreed to keep the old colouring rather than the really dull ones. The example provided shows almost 3200 hours of sunshine, which is abundant for the vast majority of the world and should really have a much brighter colour. ---华钢琴49 (TALK) 14:33, 11 June 2010 (UTC)


 * Temperatures: the red was moved later per Eduardo Sellan III opinions at the beginning of this discussion, it wasn't moved the full ten degrees he suggested though. Precipitation: if the green was one option then it should include snow too, but I don't think green snow is a good idea, its already written into the code as scsnow, but its not mentioned on the documentation. Sunshine: I thought I had it close to what it was before, but I'll try again. 117Avenue (talk) 23:42, 11 June 2010 (UTC)


 * I agree that the very idea of 'green snow' is a poor idea. for sunshine, I commend you for apparently fixing the monthly amount in order to reflect daily amount as well. all I ask for is if you could somehow change the annual amount back to what it was for the reasons I gave above. and let's not ever colour percentages (shouldn't have to explain why). ---华钢琴49 (TALK) 00:14, 12 June 2010 (UTC)
 * How's that? 117Avenue (talk) 08:28, 12 June 2010 (UTC)
 * the upper range (>3000) is better, but I think the mid-upper range (2600-2999) is still a little dull, and noticeably duller than what it used to be (see Beijing). 华钢琴49 (TALK) 18:36, 12 June 2010 (UTC)
 * Its actually brighter now then it was originally, 90 (1096/yr) was #9B9471, now its #AAAAAA.


 * ok. now I am very much satisfied with the sunshine colouring. I guess the issue may be with my laptop and/or the angle I view it from? anyway, bear in mind that Eduardo Sellan III is from Brasil, and with no offence intended to natives of that land, their standards as to heat and cold are probably higher than the global average -_- (i.e. more difficult to feel hot and easier to feel chilly). I made a mistake interpreting what is the beginning of red. It appears that 35 C is the new threshold, and I will be somewhat satisfied if the red is moved down to 33 or 32 C. At the lower end, the light blue should represent freezing temperatures. The new colouring for 1 C is ridiculous, in that it makes the tables of Tokyo, Shanghai, Vancouver, Atlanta, and countless other cities with January lows in the 0 to 1.9 C range have this blue tint, making the nights seem colder than they

really are, when the truth is that their winters are mild. ---华钢琴49 (TALK) 05:19, 28 June 2010 (UTC)


 * change my mind regarding 30 C. again, appears to be an issue with screen imaging and/or the angle. ---华钢琴49 (TALK) 22:58, 28 June 2010 (UTC)

Documentation example
ok. I understand 117Avenue's reasoning that Iqaluit is too polar and damp. however, is snow the norm for a large percentage of the world's population? is the consistent 24 C day-night difference typical of the world? both of these answers are no. so let's not fabricate data that is obviously not possible in real life: not only is the diurnal range unthinkable, but the high amount of summer sunshine despite the frequency of precipitation is questionable. However, I understand the difficulties of supplying actual data, since only Canada uses humidex, and few meteorological agencies provide both precipitation, rain, and snow data. ---华钢琴49 (TALK) 00:20, 12 June 2010 (UTC)
 * Does the example need to be something possible in the real world? I was attempting to appease all regions, there are places with a record high higher than 35, and places with a record low lower than -24, where precipitation is more than 210, and zero for months, where it rains everyday, or not at all, where it is sunny all the time, or dark for a month. 117Avenue (talk) 08:28, 12 June 2010 (UTC)
 * ok then I understand your reasoning. I am too serious at times xP ---华钢琴49 (TALK) 16:21, 12 June 2010 (UTC)

°F to °C and °C to °F precision
Once precision for temperatures in °F reaches 0.1 °F, the Celsius conversion is to the nearest hundredth of a degree. Annual averages, even more absurdly, are to the nearest three decimal places. I ask: is this necessary, and is it aesthetic? if no one objects, I will revert the °F-°C conversion to what it was before June 11 (i.e. preceding 117Avenue's massive changes). ---华钢琴49 (TALK) 18:31, 12 June 2010 (UTC)


 * Sorry, still a few kinks to work out. My intention was to add a decimal when moving from F to C, because celsius is smaller than Fahrenheit, but remain the same from C to F, then add a decimal only for automatic averaging. 117Avenue (talk) 04:15, 13 June 2010 (UTC)

what I really want for F-C monthly values is one decimal place no matter what. Two places for the annual value should be fine, but preferably no more than that: temperatures are rarely reported to the nearest 0.001 C anyways. ---华钢琴49 (TALK) 13:15, 14 June 2010 (UTC)


 * While its true measurements aren't taken to the thousandth of a degree, these are averages from thirty years or more, so the precision is possible. 117Avenue (talk) 14:22, 14 June 2010 (UTC)

Sometimes, if there is a subzero temp on the °C scale, and if the ".0" is added, then the conversion to °F fails to add the extra decimal if necessary, i.e. -6.0 °C = 21 °F (actually 21.2 °F) ---华钢琴49 (TALK) 20:30, 22 June 2010 (UTC)


 * It looks like the template, which uses parser functions, can't understand the dash you're using. Yet another reason why the parser functions aren't set up to use Wikipedia's dash MOS. 117Avenue (talk) 00:13, 23 June 2010 (UTC)

I've just noticed this template. I don't like the mandatory decimal for Celsius. I'd like to see integers. Look at how the BBC shows integer Celsius for Rome: http://www.bbc.co.uk/weather/world/city_guides/results.shtml?tt=TT004000 That is how I'd like it. Furthermore, wouldn't it be easier if this template used the well known convert template ({convert|10|F|C})? Then the code would be easier to understand and maintain. Lightmouse (talk) 20:18, 1 August 2010 (UTC)


 * What do you mean mandatory decimals for Celsius? Decimals aren't mandatory, you an use as much precision as you like. And no, the convert template wouldn't work because it is needed for the colours, the units take up more space, and you can't use line break in it. 117Avenue (talk) 21:07, 1 August 2010 (UTC)

With respect of my phrase 'mandatory decimals', I read your text above: I took that to mean that I can't have integers for both F and C. If I can have any precision, could you tell me how? Lightmouse (talk) 21:18, 1 August 2010 (UTC)
 * "add a decimal when moving from F to C, because celsius is smaller than Fahrenheit, but remain the same from C to F"


 * Maybe I didn't explain myself correctly, you can enter in whatever the precision of your source is, and it will adapt accordingly, (it is recommended that you do not round the source values). The reason one decimal is added when converting from fahrenheit to celsius is a concern that was brought up earlier. If in one month the temperature was 42°F, and 43°F in the next, the celsius row would display 6°C for both, giving the false impression that the two months are the same. 117Avenue (talk) 00:05, 4 August 2010 (UTC)

It isn't a false impression that the months are the same, it's an inherent feature of precision. The temperature one month could be 42.4 °F and 42.5 °F the next so are only 0.1 °F apart yet we accept the 'false' impression that they are 1 °F apart. Precision isn't always an arithmetic calculation, it is often a matter of rounding to integers because that is often convention by the the target users. Americans frequently quote temperatures in integer Fahrenheit, just as non-Americans frequently quote temperatures in integer Celsius ("the temperature reaches 43.3 °C in summer" rather than "the temperature reaches 43 °C in summer"). The convert template copes with this issue very well by allowing the user to choose rounding, and the default is integers ({convert|110|F|C} --> 110 °F (43 °C)). Can you make the extra decimal place in this infobox a user choice rather than mandatory? Regards Lightmouse (talk) 15:57, 4 August 2010 (UTC)

Replace hyphens with minus signs
Could someone please replace the hyphens with minus signs in this template, to make the typography correct? --bender235 (talk) 10:32, 13 June 2010 (UTC)
 * As explained above, it is a parser issue, not this template. 117Avenue (talk) 21:47, 13 June 2010 (UTC)

Year column --- Temps
Several issues here, mostly with the temperatures. You may have noticed them already, but anyways... : Thx ---华钢琴49 (TALK) 01:45, 14 June 2010 (UTC)
 * When inputting with Fahrenheit, and removing the annual record high/low parameter, the max/min calculation shows but not the conversion and subsequent colouring.
 * Averaging normal temperatures 'creates' an extra decimal point in the annual value.
 * I have noticed the problem with the automatic max and min, and I am hoping its just server lag, because of all the templates and expressions inside templates and expressions. Strangely enough if you add subst: at the beginning of the template, then click show preview, there is no error, but it is not recommend that you substitute this template, the only thing I can do is wait. As I mentioned above, the extra decimal when using automatic averaging is intentional, it is possible because the sum is being divided by more than ten. 117Avenue (talk) 05:41, 14 June 2010 (UTC)
 * actually I change my mind regarding the two decimal places, because sometimes monthly averages are reported to the nearest 0.01 F or 0.01 C. And WRT the SUBST: template, I don't know why I would bother using it and going to the trouble of doing so. ---华钢琴49 (TALK) 13:19, 14 June 2010 (UTC)

Mean Number of Precipitation/Rain Days
On page of World Meteorological Organization at UN is written "Attention: Please note that the averaging period for climatological information and the definition of "Mean Number of Precipitation/Rain Days" quoted in this web site may be different for different countries. Hence, care should be taken when city climatologies are compared". I agree with that. Data from Meteorological Agencies are different, some use 0.1mm, others 1mm and Americans - inches. Therefore, to the infobox should add the new option, what size was used, example |Precipitation/Rain Days unit =<-Here write the parameter (0.1mm or 1.0mm or other) (or any other name of option). Result, rather than "Avg. precipitation days 17 17 15 17 16 10 8 9 11 16 17 18 171" should be example "Avg. precipitation days (for 1mm) 17 17 15 17 16 10 8 9 11 16 17 18 171". Subtropical-man (talk) 15:04, 19 June 2010 (UTC)


 * 'bout time you brought this up here, so thanks. I also mentioned this issue at the Meteorology WikiProject talk page. ---华钢琴49 (TALK) 16:55, 19 June 2010 (UTC)
 * I don't understand what he is trying to say with his bad english. Are you talking about the minimum precipitation requirement for a day to be called a rain day, or the unit input for precipitation amounts? 117Avenue (talk) 19:05, 19 June 2010 (UTC)


 * Please, read: "Attention: Please note that the averaging period for climatological information and the definition of "Mean Number of Precipitation/Rain Days" quoted in this web site may be different for different countries. Hence, care should be taken when city climatologies are compared" (text by World Meteorological Organization at UN). Examples from Wikipedia: Los Angeles, California - Avg. precipitation days = 35.2 at the parameter = 0.254 mm (0.01 inches); Faro, Portugal - Avg. precipitation days = 90 at the parameter = 0.1 mm; Barcelona, Spain - Avg. precipitation days = 55 at the parameter = 1.0 mm. Wikipedia introduces disinformation. Now, do you understand? This is a serious matter. Then determine which parameter (0.1, 1.0 or 0.254mm) is used by Avg. precipitation days = on the infobox. Subtropical-man (talk) 20:11, 19 June 2010 (UTC)
 * Okay you're talking about the minimum requirement of rain for a day to be called a rain day. 117Avenue (talk) 22:02, 19 June 2010 (UTC)


 * There are three where the definition would be required, rainy, snowy and precipitation days. Would it be possible to have another line where the amount required could be added in the format "required amount=" and then the number. Then when the infobox is displayed it would show "Avg. rainy days (=> 0.01 mm)" or "Avg. rainy days (=> 0.01 in)". The mm/in would be based on the metric_first line. I tried adding the Ref to the city examples above but it didn't work correctly. Enter CBW, waits for audience applause, not a sausage. 22:50, 19 June 2010 (UTC)

How's this?
 * --- 117Avenue (talk) 00:06, 22 June 2010 (UTC)
 * obviously a little fat and there is no such precip measurement as small as 0.01 mm, but otherwise perfect. ---华钢琴49 (TALK) 00:29, 22 June 2010 (UTC)
 * Sorry, you're right, I was basing it off of CBW's example. 117Avenue (talk) 00:48, 22 June 2010 (UTC)
 * Looks good. Sorry about that I put the decimal in the wrong place and used the wrong number. It should have been 0.2 mm, 0.2 mm precip and 0.2 cm snow. Enter CBW, waits for audience applause, not a sausage. 03:06, 22 June 2010 (UTC)
 * Appearance - ok :), now need to enter this option to Template:Infobox weather with the possibility to manually input |name_parameter = 0.1mm (or 1.0mm or 0.254). Subtropical-man (talk) 17:24, 22 June 2010 (UTC)
 * ✅ 117Avenue (talk) 05:37, 23 June 2010 (UTC)


 * It works :) Is OK. Thank you 117Avenue :) Sorry,  I do not want to be "nuisance" - not better without   ?:

Subtropical-man (talk) 16:46, 23 June 2010 (UTC)


 * whatever. anyway, I am going to delete those , as the infobox seems to be expanded with their inclusion. ---华钢琴49 (TALK) 20:24, 23 June 2010 (UTC)
 * That will make average precipitation days the widest text in the first column, which may cause some problems on thinner browser windows, but NoWrap should fix any weird wrapping. 117Avenue (talk) 23:39, 23 June 2010 (UTC)

Green precipitation and rain
Subtropical-man has objected to my change of the infoboxes of many cities to display green colouring for rain or precipitation. My rationale is that:
 * I doubt 117Avenue would have not included a parameter if he believed it was "unnatural" (green snow is unnatural) or would incite mass opposition. Am I correct? I hope I did not misgauge you.
 * Green is not any more unnatural for rain and total precipitation than blue is: Rain and snow are both colourless. "Green rain", in my opinion, is better used in conjunction with warmer climates (threshold that I do not yet, perhaps -3.0 C). ---华钢琴49 (TALK) 18:45, 26 June 2010 (UTC)
 * It was good, now it looks non-natural. Rain is colourless, but has a blue tint. Blue colour is better and this is standard on Wikipedia. If you want to change on a large scale, should be a consensus of between users. Now, let's start a discussion to develop a consensus. Subtropical-man (talk) 18:53, 26 June 2010 (UTC)
 * discussion is less useful if you merely repeat old arguments. I think the problem may well reside with the wetter range of the green. ---华钢琴49 (TALK) 21:28, 26 June 2010 (UTC)
 * I do think that the colour green is natural (rain makes plants green), and this isn't the first alternate colouring made available, the temperatures can be made pastel. However, I think that it may be wikipedia policy for all cities to use the same templates, so that they can be compared properly, unfortunately I don't know where to find this guideline, we may need to bring in outside opinions before an outright war takes place between which cities should be green, and which should be blue. 117Avenue (talk) 22:29, 26 June 2010 (UTC)


 * Manual of Style (infoboxes) has some information. I've left a notice there, at Wikipedia talk:WikiProject Cities and of course Wikipedia talk:WikiProject Meteorology. Enter CBW, waits for audience applause, not a sausage. 05:24, 27 June 2010 (UTC)


 * Personally, I would go for green, as it is the color used for most precipitation related products on the NWS warning map, and because I think of blue for winter type precipitation rather than rainfall. (Note that I also have somewhat of a bias, as green is my favorite color. :-P) Ks0stm (T•C•G) 05:33, 27 June 2010 (UTC)

Is this a geocultural issue? From a British perspective, I would associate rain with water with blue. When it's in the sky, it's rain; the sky gives the rain a blueish appearance (well sometimes, even in Britain!). If it has been a dry summer, the rain turns the land green, but only after it has landed and ceased to be rain. And if it has not recently been very dry, the rain has no immediate affect on the colour of the landscape, since (most of the year) the land is already green or (in winter) will remain barren. — Richardguk (talk) 12:48, 27 June 2010 (UTC)
 * PS: Lime green rain? Hideous! Torrential rainfall is more likely to lead to flooding and mudslides than to a dark green landscape. — Richardguk (talk) 12:57, 27 June 2010 (UTC)
 * I have friends from Tianjin who tell me that the sky turns somewhat greenish during the summer torrential rains. ---华钢琴49 (TALK) 20:37, 27 June 2010 (UTC)
 * furthermore, we should not compare the two colours based only on the wetter extreme. IMO, blue is better suited for the upper extreme, but green is decisively better < ~200 mm ---华钢琴49 (TALK) 20:41, 27 June 2010 (UTC)

oh my... I just realised that the green colouring sometimes looks hideous in combination with record temps, if the place has a sufficiently large range and is wet enough. See Dallas for an example? (hint hint, that's partially my editing). but often the blue colouring seems to blend with the record low row, which is not good either. ---华钢琴49 (TALK) 00:23, 29 June 2010 (UTC)

turning away from the debate for awhile: 117Avenue, could you make the scprecip and scgreen parameters accept ANY input, like the collapsed, metric first, and single line parameters do? thanks. ---华钢琴49 (TALK) 23:03, 30 June 2010 (UTC)
 * Sorry, no. It would break the precision fork templates, which use pastel as a default, and must opt in for the standard temperature colours. It is written the way it is for expansion purposes, if there were ever a need for temperature or precipitation in another colour. 117Avenue (talk) 23:53, 30 June 2010 (UTC)

accessdate
What is the accessdate parameter outside the reference for? -- Jeandré (talk), 2010-07-15t12:55z
 * The date you accessed the website of the weather service. 117Avenue (talk) 20:00, 15 July 2010 (UTC)
 * But that's already part of the reference templates, and is displayed at the bottom in the reference section. I'm asking why there is also an accessdate parameter in the infobox template itself which is shown below the table. See the italic accessdate in the article itself at Mthatha. In no other articles I've seen, are the accessdates of when the sources were last checked shown in the content sections - it should only ever be shown in the reference section.
 * Also, having a date there may cause someone to think that it has something to do with the stats (if for instance the standard normal dates aren't given with the source name), when the stats will instead be for a long period, before the accessdate. -- Jeandré (talk), 2010-07-16t10:42z, -- Jeandré (talk), 2010-07-16t10:46z

I agree with Jeandré; access date belongs as part of the reference, not part of the caption. I'm going to be WP:Bold and remove it. Jordan Brown (talk) 17:00, 25 July 2010 (UTC)

This makes  unused, so I will also remove that from the documentation. Jordan Brown (talk) 17:16, 25 July 2010 (UTC)

✅ Jordan Brown (talk) 17:22, 25 July 2010 (UTC)

Perhaps pages supplying accessdate should go into a category to be inspected to ensure that the access date is in the reference and remove it from the template invocation. Jordan Brown (talk) 17:29, 25 July 2010 (UTC)

✅ Category:Pages using deprecated parameters in Infobox weather Rich Farmbrough, 18:06, 20 August 2010 (UTC).


 * Deleted - redundant now. Rich Farmbrough, 13:38, 26 August 2010 (UTC).

Source display
Generally, articles don't mention their sources in their text; the text includes footnote markers (via references) and the actual source is shown in the references section. This template puts the source in the caption, rather than (or in addition to) the references. I would eliminate the "Source" lines at the bottom of the table, and have the sources be shown only via footnote, probably with the footnote marker in the title. I would have the "source" parameter contain the contents of the reference (caller need not supply &lt;ref&gt;), so that typically the source parameter would contain only a.

That's a big change, since it changes the semantics of the source parameter. I'm not going to do it now, and with the number of articles that use this template the transition will be tedious.

Here's a possible transition plan:
 * Add a few new parameters ref, ref1, ref2, et cetera. Their contents would be as described, the contents of a reference.  If specified, include them in the title bar.
 * Make the source lines optional.
 * Generate the "unsourced" warnings if neither refs nor sources are supplied.
 * Add pages supplying sources (rather than refs) to a category for eventual cleanup.

Thoughts? Jordan Brown (talk) 17:00, 25 July 2010 (UTC)
 * I'd say that since this is such a widely used template, don't change it. From what I've seen, the name of the weather service it put inline, then the reference is in the footnote. 117Avenue (talk) 20:03, 25 July 2010 (UTC)


 * The fact that it is a widely used template is precisely why it's appropriate to make it as clean and attractive as possible. I don't quite follow your "From what I've seen" comment; yes, that's what the template does today, but is it the right thing?  If you look at Los Angeles, for instance, you'll see that the text says quite a few things about the weather with no mention of the source of that information, except for footnotes.  I think that's the way it should be - present the information without cluttering it with source information, and keep the source information readily available in the References section. The same should be true here. Sure, it's only one line, but for two sources it's two lines, and I've got a guy over in Template:Los Angeles weatherbox who wants to put two sources into source1 to avoid the second line.  I think he's right (though I think the problem should be solved here, not there), and moreover I don't think we need even that one line of clutter. Jordan Brown (talk) 20:36, 25 July 2010 (UTC)
 * I don't see the problem with Los Angeles. It provides the name of the weather service, and it you follow the link to the bottom of the page, there is more reference information available. 117Avenue (talk) 20:54, 25 July 2010 (UTC)
 * My point is that most of Wikipedia doesn't put source material in the mainline prose, infoboxes, or tables. The factual material is presented in the mainline prose, and all source references are pushed down into the References section.  If you look at the Climate section of Los Angeles, the only sources mentioned in the section are those in the weatherbox.  Why shouldn't the weatherbox be the same as the rest, putting source information in the References section?  This would also avoid disagreements about how to handle multiple sources, since they wouldn't clutter up the table.  (Somebody had added data to the Los Angeles weatherbox from a second source, and had put that second source into the   parameter.  I moved the second source into   where Template: Infobox weather wants it, and he reverted, complaining about the second line in the table.)
 * Create a new line(s) in the Infobox weather? Why? Better to give two sources in one line. Subtropical-man (talk) 12:45, 4 August 2010 (UTC)
 * I agree, but if we want two sources in one line, the right place to fix it is here in Template:Infobox weather, not by having the calling page misuse the argument list. What I'd really prefer, though, is to get rid of the source line(s) entirely, replacing them with footnote markers. Jordan Brown (talk) 17:47, 5 August 2010 (UTC)
 * It is usual for sources cited in infoboxes (which this isn't in the normal sense) and other templates to be listed in the references section. I have changed the "citation needed" code to use the standard mechanism, for all the obvious software engineering reasons.  Sadly SmackBot is not great at dating multi-line templates, although I am working on it.  At the moment every article containing an unreferenced weather box is showing up every run. Rich Farmbrough, 16:38, 18 August 2010 (UTC).
 * Sorry, I just realized you were the one I talked to about this. My concern was that if the user hasn't input the source parameter, he's not going to input a date. 117Avenue (talk) 00:59, 19 August 2010 (UTC)
 * Anyone can add a date. This template is, however, proving tricky for SmackBot because of the degree signs in the parameter names. Rich Farmbrough, 16:00, 20 August 2010 (UTC).
 * Incidentally SmackBot is telling me this template has 353 parameters! Rich Farmbrough, 16:37, 20 August 2010 (UTC).

date parameter
I am re-adding this. If the "citation needed" functionality is to be embedded in this template, it need to support a date field in the standard way. SmackBot is now adding date fields where needed, although not without problems, as indicated above. Rich Farmbrough, 21:55, 20 August 2010 (UTC).

Sample Date Range / Commencement of Statistical Sample
I believe the chart definitely needs to be clear about when the samples begin, however there appears to be no field for the commencement date of statistics. For instance, I looked up two seperate cities, Ballarat, Melbourne and Bendigo. While the maximum temperature in the article for Ballarat and Melbourne concurs with the maximum displayed on their respective graph, I found that Bendigo's does not. Because upon viewing the sources, I discovered that whereas Ballarat and Melbourne's charts were from over 100 years of data (beginning 1908), Bendigo's sample (beginning 1991) is less than 20 years worth of data. The variation in the duration of the statistical samples can significantly skew the reader's impression of the local climate, particularly in light of the effects of climate change and in case a particular location stops measuring data for whatever reason. --Biatch (talk) 02:34, 18 August 2010 (UTC)
 * There's another problem there. Ballarat is using 102 years of data and Melbourne is using 155 years of data for the means. They should be using the standard 30 year normals and here. The maximum and minimums should be from the entire reporting period. Enter CBW, waits for audience applause, not a sausage. 03:14, 18 August 2010 (UTC)

Proposed move to Weather box
I have created a near copy of this template at Weather box and propose moving to it with the following advantages:


 * 1) Name: This is not an infobox in the normal meaning of the word, indeed some derived templates use weather box.
 * 2) Deprecated parameters: The Category:Pages using deprecated parameters in Infobox weather shows that nearly all (at least 2/3 just with accessdate and dateformat) uses of this template need a little cleaning up.  This will be done at the same time.
 * 3) Parameter names: regularised to normal cased, spaced, no abbreviations apart from C, F and the months. Also no degree symbol.
 * 4) Citation needed: All the templates will be dated, and where an existing citation needed template follows the weather box, this date will be used.
 * 5) Blank parameters: Many cases have blank parameters contrary to the documentation, these will be removed by the process.

I have done some quick testing: the following articles seem to work fine
 * Ankara
 * Berlin
 * Marquette, Michigan
 * Avignon
 * Badajoz
 * La Rochelle
 * Tombstone, Arizona
 * Indian Wells, California
 * Escondido, California
 * Macon, Georgia
 * Lynn, Massachusetts
 * Bozeman, Montana
 * Stroudsburg, Pennsylvania
 * Inverness
 * Independence, Missouri
 * Iwo Jima
 * Mersa Matruh
 * Santa Ana, El Salvador
 * Lüderitz
 * Constantine, Algeria
 * Ouarzazate
 * Rowville, Victoria
 * Blackmans Bay, Tasmania
 * Dinajpur District (Bangladesh)
 * Abha
 * Tamanrasset
 * Geography of Toronto
 * Cerro El Pital

Comments? Rich Farmbrough, 17:04, 22 August 2010 (UTC).

There's a problem somewhere:

The first is from Cambridge Bay, Nunavut, the second is from Toronto weatherbox and the third from Yellowknife. Enter CBW, waits for audience applause, not a sausage. 18:08, 22 August 2010 (UTC)

They should be converted as above, using "Subst:Weather box maker". Rich Farmbrough, 19:24, 22 August 2010 (UTC).
 * Looks OK? Rich Farmbrough, 19:30, 22 August 2010 (UTC).
 * OK that works now. Is the updating something that a bot can do? Enter CBW, waits for audience applause, not a sausage. 22:56, 22 August 2010 (UTC)
 * Yes I can AWB it either manually or botwise. So far I have been doing it manually and checking each one. Rich Farmbrough, 13:21, 23 August 2010 (UTC).
 * As the creator of "Infobox Weather" I say: NO, NO, NO, NO!


 * Ok, Now that I have your attention [and stop rolling your eyes], I am/was just joking. I am glad someone can replace some of bubblegum, bobbypins, and duct-tape that I may have used to piece this thing together with nuts and bolts (so to speak).  But remember to deprecate this template and after all the AWB switches, merge all the old "Infobox Weather" talk pages over to "Weather box talk" and have Template:Infobox weather redirect to "Weather box".  Good job and good luck.  &mdash;  MJC detroit  (yak) 02:20, 24 August 2010 (UTC)
 * Thanks for those good wishes, there is an amount of fiddling about to be done after migration, and as you say, some improvements needed. I will try and do as much of the tidying up as possible. Rich Farmbrough, 00:06, 25 August 2010 (UTC).


 * now the code is far better and more sensible. no more exceedingly obnoxious underscores! Issues are that humidity and green precipitation/rain do not show, even after deleting underscores. But I understand that that is what occurs after a move, and shall thank you for the rest. ---何献龙4993 (talk) 19:06, 26 August 2010 (UTC)