Template talk:Birth date

Non-Gregorian dates
I have added an error message to the sandbox version that generates a message when the template is incorrectly used for a non-Gregorian date. If there is consensus for this change, I will ask a template editor to make the change. Celia Homeford (talk) 13:35, 7 September 2018 (UTC)


 * Wherever this change is documented, it should mention that while a year earlier than 1582 guarantees the microformat should not be emitted, and thus the template should not be used, later dates might also be Julian, and if so, the template should not be used. Jc3s5h (talk) 15:42, 7 September 2018 (UTC)

Please update the template from the sandbox version to implement the above change. Thank you. Celia Homeford (talk) 08:25, 4 October 2018 (UTC)
 * Sorry to be raising questions at this stage but the sandbox has code in it to check for "YYYY-MM-DD". What is that? Was the sandbox copied from the main template before editing? Also, if there is code to show an error if year < 1582, why not just omit the microformat instead? That would not solve the problem noted by Jc3s5h above, but it would be useful. This link shows the diff between the current main and sandbox templates. Johnuniq (talk) 09:55, 4 October 2018 (UTC)
 * I've removed the trim code per Template talk:Birth date/Archive 2. As I understand it, removing the microformat will disable the ability of other websites to automatically read wikipedia (see Help:Microformats). I will ping User:Pigsonthewing as he is the Founder of WikiProject Microformats. Celia Homeford (talk) 10:20, 4 October 2018 (UTC)
 * What I meant was to include the microformat if the year is 1582 or later, but omit it otherwise. Johnuniq (talk) 10:44, 4 October 2018 (UTC)
 * This has been discussed before; see Template talk:Birth date/Archive 2 passim. My takeaway from that: a) the Gregorian calendar was not universally introduced in 1582; basing this template's behaviour only on that simple numerical comparison will have unintended consequences. b) This template's documentation already advises against its use when inappropriate; this can only be decided in each case. c) The sky is not going to fall in when this template is mistakenly used inappropriately. Summary: It's probably better to do nothing here; when inappropriate usage is found in articles, the date should be rewritten raw, without the template. -- Michael Bednarek (talk) 11:00, 4 October 2018 (UTC)
 * But dates before 1582 are always wrong. I see only benefits from easily identifying articles where we know the date is emitted incorrectly. Celia Homeford (talk) 11:06, 4 October 2018 (UTC)
 * A copy of the latest draft of ISO 8601, which I obtained while it was posted at the Library of Congress website (but since removed), which claims to agree with the current 2004 version of the standard in the section of interest, stated "Values in the range [0000] through [1582] shall only be used by mutual agreement of the partners in information interchange." Since the hCard microformats follow the ISO 8601 standard, and since the described mutual agreement does not exist, years less than 1583 are always incorrect. The criterion year < 1582 is the wrong criterion for the suppression of the microformat; the year should be 1583. Jc3s5h (talk) 13:00, 4 October 2018 (UTC)
 * OK. Celia Homeford (talk) 13:09, 4 October 2018 (UTC)

I have a dump of the wikitext from all articles as at 1 August 2018. Searching that for uses of birth date gives these results: 1,650 articles have birth years from 32 (Otho) to 1582 inclusive (14 of these, for example Taichang Emperor). That means this template must not simply display an error if year < 1583 because that would create a must-fix-now list of 1,650 articles. Instead, the hidden tracking category could be added and we could encourage wikignomes to "fix" the articles over the next couple of months. Fixing by removing the template would be sure to generate trouble because others won't understand why it is being removed, and if they were given an explanation many might resist because using for a birth date just seems natural—if there is a problem, the template should solve it (they would think). I'm not sure what would be best but suspect that leaving things as they are would be the right outcome. By the way, while sorting the dates from the dump I noticed many broken templates, for example February 20, 1943 here and April 4 here. If I converted this to use Module:Age (like birth date and age) invalid input would give an error. Johnuniq (talk) 10:12, 6 October 2018 (UTC) I think we will have to educate lots of users anyway. There will be more articles with dates precise to the day for people who were born more recently. Also, since this is the English Wikipedia, there will be a disproportionate number of articles about people from the UK and US, in the period beginning 1583 and ending 1752, where the Julian calendar was in effect in the UK including parts of the future US. For these articles we won't be able to rely on just supressing the metadata. Jc3s5h (talk) 15:28, 6 October 2018 (UTC)
 * Yes, that's true. What do you think about my feeling that removing the template would lead to problems such as the perpetual urge by wikignomes to add what they might think was a missing template when they find the article? Perhaps there could be a parameter to cause suppression of the metadata; that would make the edit self-documenting. Johnuniq (talk) 00:30, 7 October 2018 (UTC)
 * When there is not appropriate to create the metadata, the template serves no purpose. I'm not enthusiastic about adding worthless templates, with an extra template to boot, when the template is useless. To bad we can't create a system like that used at WikiTree which requires users to pass a quiz before they are able to enter old dates. Maybe it would suffice to create a Userbox that would allow people to brag they passed the old dates quiz. Jc3s5h (talk) 12:31, 7 October 2018 (UTC)

Request disable for now. If reactivating, please clarify which changes have consensus. &mdash; Martin (MSGJ · talk) 21:30, 10 October 2018 (UTC)

I've done about 95% of the 1650 and amended the sandbox to just add the tracking category. Celia Homeford (talk) 09:51, 18 October 2018 (UTC)