User talk:Rkwiki

Dealing with implausible and unsourced claims
Hello, Rkwiki. I just noticed your edit summary at Jantar Mantar. Be bold! If I'd noticed it, I would have removed it, and explained my action in an edit summary. I'm now having second thoughts about the capacity of a gigantic sundial to show local time within 2 seconds accuracy; it might be possible, but it would require quite a challenging set of calculations. I've found no reliable scientific source by a reputable author or publisher to confirm it. The sources making the claim in the article are inadequate, and should not be used. Haploidavey (talk) 18:07, 18 April 2021 (UTC) Hi Haploidavey, I do not think that a sundial can be read to a few seconds. But there are other wikipedia pages that make the same claim, that the Jantar Mantar sundial can be read with two second accuracy, so I thought I would go through them one at a time. Rkwiki (talk) 23:06, 18 April 2021 (UTC)rkwiki


 * Just to follow this up; the accuracy of a sundial depends on its size, and the user's ability to calibrate a noon reading. The longer the gnomon, the more accurate the reading can be. As long as the sun shines! Haploidavey (talk) 04:12, 19 April 2021 (UTC)


 * This source seems scientific, reliable and definitely useable for Wikipedia: see p. 208 - https://www.ias.ac.in/article/fulltext/reso/022/03/0201-0212 - "the instrument is intrinsically capable of measuring time accurate to a second!" Note the "intrinsically capable." Settlement and slight erosions of the structure have negatively affected its actual and current accuracy; the original levels of accuracy could be restored. I'll replace the horrible source and quote the new one. Haploidavey (talk) 05:30, 19 April 2021 (UTC)

Really, this question about accuracy should not be difficult to answer. Have someone go there with a cell phone which uses time from a local carrier that has been checked to be accurate. Take closeup pictures of the dial that show the edge of the shadow (the umbra) and the 2 second, six second, and minute marks (if possible!), and a few wide angle pictures that show the hour and 15 minute marks so that we know what part of the dial was being checked. Then put the pictures online with the EXIF available, and we will be able to see if the suntime on the dial plus the EOT correction for the date matches the time the photos were taken. I have done this myself with a nearby large horizontal sundial, and it is not difficult. A few caveats, the person taking the pictures of the sun time should not be looking at the clock time and does not need to know the EOT correction for that date. Do this on a few different dates and times, and then I will be convinced of the accuracy of the sundial! If someone will do this, I will donate the admission fees. There should be someone nearby willing to do this. Thanks! Rkwiki (talk) 06:49, 19 April 2021 (UTC)rkwiki Hi, update on my last comment. The Wikipedia article we are talking about https://en.wikipedia.org/wiki/Jantar_Mantar which I corrected to "The Vrihat Samrat Yantra is a sundial that is claimed to give the local time to an accuracy of 2 seconds.[2]" is concerned with the sundial at Jaipur. That is not clear from the Wikipedia text, but reference [2], which makes that claim, is about Jantar Mantar Jaipur. However, in the article you link to https://www.ias.ac.in/article/fulltext/reso/022/03/0201-0212 the calibration September-December, 2006, was at the _Delhi_ Samrat Yantra (page 206). The readings were conducted "on winter solstice day 2006" (figure 7, page 209). Ok, whatever, but the winter solstice is an odd day to pick to read a sundial, as when the sun is low in the sky the sunlight has to go through more atmosphere which causes more diffraction. And the azimuth of the sun moves a shorter distance every hour, which puts the time readings closer together. And it is odd that the article doesn't show photographic evidence of the readings. Anyway, now with cell phones with good cameras and accurate time, it should be easy to document the accuracy of these sundials, so I will wait for that.Rkwiki (talk) 17:47, 19 April 2021 (UTC)rkwiki


 * Kudos for your attitudes here. I've a few committments this evening, and will have to come back to you (and the several issues arising inthe matter of accuracy) tomorrow. Meantime, you do realise (?) that any experimentation-based research you do on this will count as the dreaded WP:OR. So unless you're going down that route simply to satisfy your own curiosity, or to confirm or otherwise comment on the linked research, none of that material will be usable in Wikipedia. It might also be wise to recall that Wikipedia doesn't claim to sell the truth, nor even accuracy. To put it another way, the results of the linked research may well NOT be too inaccurate to use. It doesn't claim to be right, nor should it. It doesn't need to be right, correct or what have you. Just accurate enough to justify an approximate range, derived from standard statistical methods and distribution analysis. It might even be possible to contact the authors (or at least one of the main authors), to clear up the doubts regarding methodology. Haploidavey (talk) 18:08, 19 April 2021 (UTC)

In the article you linked to, the data was collected in 2006. Very odd (to me) that it was published without photos illustrating their readings. Anyway, I don't think I will try to contact the authors 15 years later, when now, as I said, a modern cell phone could do a good job of checking accuracy. The link [2] in the Wikipedia article is a .php that dates from 2016, but it is not a scientific article.Rkwiki (talk) 06:34, 20 April 2021 (UTC)rkwiki


 * Actually, the publication is described as a general article (aimed at a general reader). My guess is that they'd have edited out a lot of evidential material that attended the original submission; or, more likely, asked the author to meet a required maximum word-count, which would have included any appendices. I bet the work's been published in more than one journal or is held in a research library, in more detailed formHaploidavey (talk) 18:27, 19 April 2021 (UTC)

Resonance is a journal of the Indian Academy of Sciences, published since 1996. The article about Jantar Mantar was published in volume 22, March, 2017, although the study was performed in 2006. Author: N Rasthnasree, Nehru Planetarium, Teen Murti House, New Delhi, her email address is given, and a link https://astron-soc.in/outreach/activities/archaeo-astronomy-projects/observing-at-the-jantar-mantar/ The only photo of the Delhi Samrat Yantra in the article is Figure 6, which was taken in 2005, before the new markings were put in place in 2006, and the markings on the dial cannot be seen even when Figure 6 is enlarged. Figure 7 is a blurry graph, but when the pdf is enlarged it can be seen that most of the observations are claimed to be accurate to 1, 2, or 3 seconds. 425 (or 450 according to the article text) observations in six hours, more readings during peak hours, so 1-2 readings per minute. Did they have two people working independently (one to read the sundial, and the other to read the clock and record the results)? This is a 12 page article, so there could have been room for details about how the study was performed. Science needs to be reproducible, https://en.wikipedia.org/wiki/Reproducibility If someone thinks that a sundial could be read to the nearest few seconds (I don't) they should do another study to document this. If these sundials really are that accurate, why do you not find cell phone photos online, with EXIF, demonstrating that? (Rhetorical question) Sorry, but I think there is something odd about these results. Rkwiki (talk) 06:34, 20 April 2021 (UTC)rkwiki