Wikipedia talk:WikiProject National Register of Historic Places

Wikipedia:Wikipedia Signpost/WikiProject used Wikipedia:Wikipedia Signpost/WikiProject used

Some thoughts on Wiki Loves Monuments

The Wiki Loves Monuments (WLM) photo contest goes Sept.1-30 and we are not part of it this year. It's fine that we aren't if we aren't going to organize WLM-US for our needs, but this might be a good time to do some review and see if we want to organize something for next year. A short table should do it using some approximate numbers.

Year	Total Photos from contest	%Illustrated 8-31	%Illustrated 10-6	Notes
2012	20,000	44.6%	49.9%	1st year
2013	10,000	57.1%	60.4%	run by Wikimedia DC
2014	10,000	65.5%	68.4%	"Summer of Monuments" not WLM
2015		72.3%	73.3%	no contest

The % illustrated is the % of NRHP sites with at least one pic, Aug 31 is the day before WLM Oct 6 is the date when most WLM pix should be included in the tables (also 36 days after Aug. 31 = 1/10th year)

The first thing to note is that the % illustrated went up by 27.7% in those 3 years, 11.5% was in the WLM period (Sept), 16.2% during the rest of the year. This indicates that September is a very important period (though not all uploads in Sept are WLM). I think probably more important than the gain in September is that several photographers came to the project during WLM and stayed to contribute throughout the year. In that sense, I think we can say that WLM has contributed about half the gain in % illustrated over the last 3 years

The next thing I notice is that we are likely to be at about 79% illustrated at this time next year if we get a 6.8% gain like we did last year. It would be nice to have a WLM next year to help get over the 80% mark. Of course, the closer we get to 100% illustrated, the harder it is to make progress. The unillustrated sites are going to be the ones way out in the boonies, on closed military bases, etc. I wouldn't see much point in having a WLM-US after we get to 80% - we'd just be attracting people to take pix of sites previously illustrated. So next year might be the last year it makes any sense for us to have a WLM.

I would very much like to have 1 last WLM next year, but carefully tailor it to our needs. That will take some work and planning. More on this later. Smallbones_(smalltalk) 19:11, 26 August 2015 (UTC)

Some of the things we might want to do to tailor the contest would be to have special prizes for the 6 states now below 50% illustrated. We might want to contact state historical societies in those states to help publicize the contest and to donate archival photos. (That's pretty much what Summer of Monuments tried)

We could also have special prizes for the most previously unillustrated sites photographed, get some "edit-a-thons" in those states - not to edit, but on finding sites, reading our tables, and uploading pix.

Other ideas would certainly be appreciated. Smallbones_(smalltalk) 19:57, 26 August 2015 (UTC)

In the boonies? I literally live near the Boone Docks, see File:Boone Dock Road sign in Darien, GA, US.jpg. :-) Bubba73 ^{You talkin' to me?} 19:45, 26 August 2015 (UTC)

I knew there was something special about you!  :-p All the best. Smallbones_(smalltalk) 19:57, 26 August 2015 (UTC)

I've added your pic to the collection of my favorites at User:Smallbones#Positive signs

The sign is in Darien, Georgia. I live about half way between Darien and Brunswick. Bubba73 ^{You talkin' to me?} 15:38, 27 August 2015 (UTC)

Some things that would be nice to have before the 2016 WLM:

In the area of emphasizing that we're looking for photos of currently non-illustrated items...

• Pages/categories of non-illustrated listings
• Something to click on the county list tables to only show non-illustrated listings
• A mod to GeoGroup so that, on the county list pages, it also offers to map only the non-illustrated listings
• Get the Map Tool back up, with a way to only show non-illustrated listings
• Some way to modulate the non-illustrated listings to note address restricted items
• A WLM guide to finding properties, with tips on trip preparation and dealing with the sometimes inaccurate locations in nris and the nomination forms.

And, of course, ErfgoedBot. Generic1139 (talk) 15:03, 27 August 2015 (UTC)

ErfgoedBot is working again! So is Wikipedia:WikiProject National Register of Historic Places/Unused images with 140+ images. Smallbones_(smalltalk) 04:05, 1 September 2015 (UTC)

There is also a working map at https://tools.wmflabs.org/wlm-maps/#2/0.2/-0.2

Smallbones_(smalltalk) 14:06, 1 September 2015 (UTC)

Thanks for the map pointer. It isn't working at the moment (loading, please wait) but previous iterations also had such transient issues with fetching monument data. Generic1139 (talk) 15:29, 2 September 2015 (UTC)

"Of course, the closer we get to 100% illustrated, the harder it is to make progress. The unillustrated sites are going to be the ones way out in the boonies, on closed military bases, etc. I wouldn't see much point in having a WLM-US after we get to 80%." There are certainly going to be ones we can't get, but I think that is well under 20%. Bubba73 ^{You talkin' to me?} 03:59, 5 October 2015 (UTC)

break

I've put in the number 73.1%+ for Oct. 6, 2015 (now updated to 73.3%) even though we're not quite through with September yet. This is to emphasize that we'll likely get less than a 1% increase during the 36 day period this year, compared to a worst result of 2.9% when we had some sort of WLM/SoM contest before. As I said above, I think that next year will be the last time it will make sense to have a WLM contest in the US. I propose that we do the following:

• contact @Harej: and the DC chapter to see what they can do to help out for next year, possibly including:
  • coming up with some prize money
  • hiring a part-timer to contact state and local historical organizations who might have photos to upload
• doing many of the things suggested by @Generic1139: above
• getting a planning committee of at least 3 people to work a few things out
• seeing how many folks would be willing to tentatively commit to working about 5 hours per week on this next September. I'd have to limit myself to about 10 hours per week, but don't see much point unless we can sign up at least 5 editors willing to (tentatively) commit 5 hours per week. I'll ping the obvious folks @Nyttend, Dudemanfellabra, Daniel Case, TheCatalyst31, Bubba73, and Magicpiano:, but anybody else who's interested should just note that below.

Smallbones_(smalltalk) 19:41, 30 September 2015 (UTC)

• I don't think I should commit myself to anything for next year yet; it all depends on where I'm living and what I'm doing by that point, since I'm currently looking for work all over the place and don't know what I'll be able to get. I'd be interested in helping, of course. Now aside from that, I'm particularly interested in the idea of presenting people with maps of currently unillustrated sites. If we could do that, it should draw additional attention to those sites (especially useful for people in the under-50% states) and reduce the number of new images from heavily illustrated spots like the Statue of Liberty. If I encounter WLM for the first time and discover a huge list of places that might or might not need images, where am I going to start? Are they just trying to get images for unillustrated spots, or better images for illustrated spots too, or something else? But if I find that WLM people have emphasized just a subset of that list, I'll have a better idea of how I can help, especially once I discover that the subset is the unillustrated spots. Nyttend (talk) 20:16, 30 September 2015 (UTC)

@Smallbones:I'd be happy to pledge some time next September (and presumably some would be required before that). As regards last year's contest, it was lacking a bit in follow up, only the top three winners were contacted; I don't think any of the other finalists were, and there was a five month lag between the end of the contest and announcing the winners. Yes, it is all volunteer, but we need to do better next time. I put my 3rd place winnings back into the pot, but some barnstar or other recognition would have been nice for the finalists. The results were posted in March. The project results in the form of a grant report were also posted. One outcome was that there was not much in the way of institutional uploading by March 2015.

If we are going to emphasize uploading images for articles that aren't currently illustrated, I wonder if cash prizes are appropriate - while there are notable gaps, some areas of the country are mined out. One other group task - now that there are more nomination forms available, and we've tagged listings without locations, a pre-September effort to reduce the number of such listings is in order.

Another pre-September thing we would could do is prepackage some trips- I always make a document with capsule description of what and where the target is, and vett the location with google/bing maps if possible. My first year of taking NRHP pics included much frustration as I learned that the locations were not accurate and learned how to actually find these things. Pre-trip prep allows for a much higher success rate.

Generic1139 (talk) 20:40, 30 September 2015 (UTC)

I might be willing to devote some time, but I don't know much about the project or what I'd need to do. Bubba73 ^{You talkin' to me?} 01:19, 1 October 2015 (UTC)

Something that would probably be useful both for photo hunters and for us if we make any maps, and that we could start as soon as we want, would be to preemptively check the locations of some of the unillustrated sites and correct the bad location info. I'm not sure we could get through all 25000 unillustrated sites, but we could at least go through some of the larger unillustrated areas (states/counties below a certain percentage, or any obvious blue pockets on the progress map). We could maybe even couple this with gathering a list of more easily-illustrated sites, so we can weed out address-restricted sites and stuff that's not easily accessible too. TheCatalyst31 ^{Reaction•Creation} 04:51, 1 October 2015 (UTC)

I don't know if I should make a "formal" commitment, but I will surely be around the entire time, and I am more than willing to help put everything together if I'm needed. I feel like there's probably a lot of automation that can be brought in to lessen the burden and speed up response times as Generic1139 pointed out was a problem last year.--Dudemanfellabra (talk) 07:00, 1 October 2015 (UTC)

I might be able to help, but I'm not much for organizing, etc. I'd rather be taking pictures, which I continue to do. I usually upload shortly after I take them - I'm not waiting until September. I have some opinions on technical and aesthetic properties of photographs I might share at a later time. But one thing that interests me is getting better photos of places. I know that any photo is better than none, but many we have are absolutely terrible. Digital cameras have changed a lot. I've even gone back to get better photos of some things I took a few years ago (I can give some examples sometime). Bubba73 ^{You talkin' to me?} 01:34, 5 October 2015 (UTC)

This looks pretty good so far. BTW how did I forget to ping @Ammodramus and Acroterion:?

It's clear that nobody can really commit to x hours per week next September, but if we have 5 editor tentatively commit then I think we'll be ok.

Some of the things we'd probably have to plan for include:

• making sure we have a system set up to help new photographers so that we can avoid all the frustrations that they can have, and so that they will stay and submit more photos later. (I think we've been fairly weak on this in earlier WLMs)
• placing the pix in both the lists and in articles (we've been great in doing this so far!)
• coordinating with the international WLM, if we want to get the benefits of working with them (e.g. the banner)
• setting up a jury to judge the final 300 or so pix (I don't think this has been a problem)
• pre-screening for photos that go to the jury (this might be the most time consuming part)
• setting up on off-Wiki website where we can blog and try to motivate photographers, thank any sponsors
• coming up with $ for prizes
• writing introductory material on "How to find NRHP sites" and "How to get better photos"
• finding an easy way to focus on nonillustrated sites and politely say "Sorry, but we really don't need more pix of the Washington Monument or the Statue of Liberty."

I'm sure there is much more, but that's a start.

I think we have at least 3 folks already who would be willing to help organize this, and who can promise to try to move things forward (until the next unexpected life event). We're close to getting another 3 who can promise some early input and labor in September (without actually committing). If we get to that point, I say we should set up a dedicated page here to do the planning.

BTW, we should ping all the founders of WP:NRHP. I'll try a few off the top of my head now @Ebyabe, Clariosophic, Denverjeffrey, Farragutful, and Pubdog:

Smallbones_(smalltalk) 17:07, 1 October 2015 (UTC)

An excellent summation of the thoughts thus far on the 2016 Wiki's NRHP project Loves Non-Illustrated Monuments initiative. I'd be happy to do a straw man for "How to get better photos", which would be a summarized version of Image guidelines, and a smattering of other guidelines, with mostly building-based examples. Generic1139 (talk) 18:32, 1 October 2015 (UTC)

@Smallbones: I can more than likely volunteer some time next September to help out, probably even the 5 hrs/week that you suggested above. While my schedule the last several years has been flexible, we're still a little less than a year out, so there is still time for things to change. To be honest I have spent many years organizing projects and I am getting to the point in my life where I am tired of it. One thing I have noted is that contributors often times do not understand how to utilize the categories. That is a task where I am more than willing to help.

A couple of thoughts about the remaining properties. I have photographed NRHP properties in places I know well and places I was passing through. The map coordinates on the Wiki listings are often times wrong, and in some cases they are way off. (I believe someone mentioned that above as well.) The National Register itself lists the wrong address for some of the properties, and are therefore wrong on the Wiki listings. I spent the last year redoing some of my photos of the properties I was not familiar with in my own hometown as I photographed the wrong building because of incorrect information. It also seems the National Park Service has recently uploaded a great many of the nomination forms and photographs of the properties on the NRHP website. That could be a great asset, but I have detected several errors in regards to the photographs. It's as though someone dropped some of the files on the floor and then tried to put them back in order. They got some right and some wrong. Another problem is that several of the properties listed, especially in the 1970s and 80s, have subsequently been torn down and yet remain on the National Register. I have noticed when going through submitted photos that some of the buildings photographed are obviously not the building listed. Recently there was a photo submitted of a modern apartment building that was described as a school. I did some research and found that the historic school building was torn down and the apartment building was built on the same property. Finally, not only are some of the properties in rural areas, but they are located in the middle of the property and in order to get them one has to trespass on private property as they are not visible from the road. I have found this to be true here in Maryland. These are a few more of the problems we are going to face going forward. Farragutful (talk) 20:20, 1 October 2015 (UTC)

I've found that simply waiting for winter has worked in several "middle of the property" cases. You are right about the addresses, but some of those are the town changing its address scheme for various reasons, including 911 rules. As you say, though, detective work is needed in some cases. I hope I have time to do some of that for regions I'll never get to this winter. Maybe we need a "lat/long verified" tag we can add to the county list pages. Generic1139 (talk) 21:19, 1 October 2015 (UTC)

I would love to help. My spare time over the next few months is likely to be limited (which is actually a good thing). However, I'll do what I can in 2016. Could we maybe create some routes using Google maps? That's how I planned my roadtrips. Yes, no GPS, me. :) --‖ Ebyabe ^talk - Welfare State ‖ 21:35, 1 October 2015 (UTC)

Since I've been on the jury for the last few years, and expect to be again, it probably would be problematic for me to get involved in any other capacity. Daniel Case (talk) 05:08, 3 October 2015 (UTC)

Well Mr. Case, you should be happy to know that I've recently added some pictures from the Lynbrook, New York, East Rockaway, New York, the Hamlet of Southold, New York, the Village of Babylon, New York, and Orient, New York. Yaphank, New York, and one last site in Emporia, Virginia should be on my list soon. For Babylon, it's the old library you mentioned. I want to put all the images into a new category, but I don't want the name to be anything generic like "Babylon Library," or "The Babylon Library." In Orient, I got the Orient Historic District, but I'm not loading any into the article, because I believe the best image for that is the obelisk at NY 25 and Village Lane, and I don't like the angle I took that picture from. I suppose I can try the Oysterponds Historical Society Building when I load images of that up, but that building is on Village Lane, not NY 25. ---------User:DanTD (talk) 15:39, 25 October 2015 (UTC)

Reference problems

Can someone experienced with fixing reference problems look at the NRHP article Frawley Ranch?

I've fixed part of the issue - not about what to do with the rest of it. --Ser Amantio di Nicolao^{Che dicono a Signa?}_{Lo dicono a Signa.} 02:22, 16 October 2015 (UTC)

Thanks

I fixed the rest of it. (Now someone just needs to expand the article...) TheCatalyst31 ^{Reaction•Creation} 03:24, 16 October 2015 (UTC)

Working on that (just a start) so thanks to everyone. I Keep running into other issues involving South Dakota and get side-tracked. If interested you can check out Talk:Wounded Knee Massacre#Problems with the Medal of Honor list recipients among other things including Talk:South Dakota#Controversial issues. The last one would seem to be something of importance. I am looking at other South Dakota NRHP articles. --- Otr500 (talk) 10:23, 21 October 2015 (UTC)

Architectural details

We discussed architectural details in the subsection "Architectural details and bricks". I recently added details to Union Banking Company Building, Obediah Barber Homestead, First African Baptist Church and Parsonage (Waycross, Georgia), Phoenix Hotel (Waycross, Georgia), United States Post Office and Courthouse (Waycross, Georgia). However, in the past, with other types of articles, several times other editors have removed photos from articles, stating some policy of not having a photograph in an article unless it is of something described in the article. That is, don't show details of a window unless it is discussed in the article. The NRHP forms usually have such descriptions, but I don't always know what they are talking about, nor do I have time to dig them out of the forms and put them in articles. Does anyone have any input on adding photos of architectural details, without them being discussed in the article? Bubba73 ^{You talkin' to me?} 17:15, 19 October 2015 (UTC)

My opinion is that the number of photos for an article should be related to the length of the article, and that, indeed, if an item isn't important enough to be featured in the article, then it doesn't need to be specifically illustrated. There are always exceptions, of course. In general though, take as many photos as are interesting, put them in a category on commons, and use one of the commons category templates to link the article to the photos. Everybody wins. Generic1139 (talk) 17:36, 19 October 2015 (UTC)

Then I'm overdoing it in those articles. These photos are some I selected out of the ones I took and I did make Commons categories for them. My feeling was that if they only exist in a commons category, and there is only the link "other images" from the NRHP county article, most likely almost no one will ever see them and the effort to take them is wasted. Bubba73 ^{You talkin' to me?} 17:47, 19 October 2015 (UTC)

(EC) I'll suggest using as many photos as needed to illustrate the building, without overwhelming the reader. The policy that is sometimes quoted is probably WP:Image use policy in particular "The purpose of an image is to increase readers' understanding of the article's subject matter, usually by directly depicting people, things, activities, and concepts described in the article." Usually is a key word here - it's not a hard and fast rule. If you have any trouble on this again, I'll suggest asking folks here on whether a removed photo depicts things, activities, and concepts described in the article. It always is a good idea to get as much info as possible out of the nomination form, but there is a lot of jargon in them, often both bureaucratic and academic jargon. It is our strength that we don't use that jargon, so if the nom talks about a window and you don't understand half of it, don't be afraid to write something as simple as "The first floor front window is considered especially interesting." And illustrate it with a photo, of course. Smallbones_(smalltalk) 17:53, 19 October 2015 (UTC)

I went over the examples you provided at the top. All of this is judgement of course and different people will have different judgements, but I'll say that those examples run the full range from having just enough photos, to having about twice as many as is needed (the surplus being mostly near-duplicates or very minor details). BTW, I love photos in articles - a picture is worth a thousand words - a couple of my own examples Newlin Mill Complex and Idlewild (Media, Pennsylvania). One of those has exactly the right number of photos, the other maybe about 50% too much. I'll let you decide which is which. Smallbones_(smalltalk) 18:14, 19 October 2015 (UTC)

I have to disagree with Bubba's statement that if a photo isn't used in an article right away, then the effort to take it is wasted.

First, by taking lots of photos, we're documenting the condition of a site. Especially if a site is deteriorating, future historians might want to know how it looked in 2015.

Second, even if I don't write an article or expand a stub, some other editor might want to do so, and that editor might be able to use my photos to illustrate specific points. As long as the photos are there at Commons, that's a possibility.

Third, Commons photos are available for uses other than Wikipedia. A photo that doesn't make it into an article might work for somebody's genealogy page or for a local historical society. Ammodramus (talk) 02:44, 20 October 2015 (UTC)

Saturday the sun was out (which has been rare for here for the last 8 months) so I went to Ware County, Georgia to the seven NRHP sites. One had no picture and the others needed better pictures. I left after lunch, took photos until I stopped for dinner, then took some more, and got home about 8PM. I took 288 photos, although some were not of NRHP sites. I think I've spent well over twice that amount of time selecting which photos to upload (I uploaded only a select fraction - well, I uploaded 53), editing them in Photoshop, giving them a good name, and uploading them to commons. I created some categories and tried to categorize the photos. Plus I created articles to have somewhere to show the photos (except for the PO which already existed and the two HDs). In all, it probably amounts to over a half hour of time for each one I uploaded. A photo that sits in the commons category Downtown Waycross Historic District may never be seen by anyone.(and I have close-up details of several of them, but I'm too tired to upload them.) Taking the photos is the fun part - the rest is a chore. Bubba73 ^{You talkin' to me?} 03:26, 20 October 2015 (UTC)

The editing and uploading and describing and categorizing can get a bit tedious. I'm looking forward to uploading my last 12 photos in Commons:Category:Auburn Historic District (Auburn, Nebraska), which'll bring the total to 81. And that one was relatively easy, because the nom form only gave approximate dates for the buildings, so I didn't have to do date categories.

That said, the photography isn't always that much fun. I spent a piece of yesterday afternoon in the hospital in Tyndall, South Dakota, getting four stitches in my hand after an attempt to photograph an NRHP bridge led to a mishap with a barbed-wire fence. Which meant, dammit, that I didn't have time to hit all the sites on my schedule, so still don't have quite enough photos to bring SD up to 30% photographed. — Ammodramus (talk) 11:56, 20 October 2015 (UTC)

Something similar happened to me last year. I was out to photograph a partial solar eclipse. I had a delicate solar filter on my lens, and it was so delicate that I had an expensive UV filter over it to protect it. I was early for the eclipse so I was walking around looking up at a building. I missed a step and my head hit the sidewalk. I had to get several stitches. It also busted my UV filter and solar filter (and I missed the eclipse). I've stumbled several times when doing other photos but haven't fallen.

I uploaded two more from my Ware County trip making a total of 55 (out of 288 taken). I could upload a total of about 250, but it takes too much time. My photos are about 10MB in size and Commons won't accept more than five or six of them or it stops processing. When that happens I have to abort and start the procedure over with a fewer number of photos. There needs to be an easy way to upload a set of photos to one commons category. For instance, if you could select a group of photos and say upload them to the Waycross Historic District category, start it, and let it go. Then it does it, and gives them generic names such as "Waycross Historic District 1", etc. Just a quick and simple way to do it - then I could upload a lot of photos to the Commons categories that would be too time-consuming otherwise. Bubba73 ^{You talkin' to me?} 16:56, 20 October 2015 (UTC)

I slipped on a rock and fell in a creek at 40.216668°N 78.182902°W / 40.216668; -78.182902 where a bridge was washed out, taking File:Minersville Coke Ovens.jpg. Managed, barely, to keep the camera over my head and out of the water. Twisted ankle, no stitches. Generic1139 (talk) 17:19, 20 October 2015 (UTC)

I was walking around this building 31.14928°N 81.49814°W / 31.14928; -81.49814 (which happens to be in Brunswick Old Town #79000727 - address restricted), not watching my step when I missed a step and hit my head on the concrete. Bubba73 ^{You talkin' to me?} 17:48, 20 October 2015 (UTC)

Watch that wire!

Of course, there had to be some irony involved in my South Dakota episode. The day before, I'd photographed the NRHP-listed veterans memorial in Winner, SD, and thought that the barbed wire by the sculpted doughboy's feet was an interesting enough detail to photograph.

I guess barbed-wire lacerations are better than getting machine-gunned by the Kaiser's minions... Ammodramus (talk) 20:14, 20 October 2015 (UTC)

My blood on the concrete

I was looking up and missed the top of these steps, and hot my head on the concrete below. That is my blood (a few days later). Bubba73 ^{You talkin' to me?} 20:25, 20 October 2015 (UTC)

Maybe we need a Purple Heart barnstar...

Regarding uploading multiple photos to a category, as you suggested above, it's possible to do that from the main "Upload a file" page in Commons. Go through the first step (picking all the photos you want to upload, and waiting for them to upload) and the second (the licensing). Once you reach the stage of writing the descriptions and doing the categorization, you can rename and describe and categorize the first photo; then there's a button like "Copy all information to files below". You can use that to automatically copy the descriptions and categories to the remaining photos. I think it also lets you copy the first photo's name to the others, with sequential numbering, but I've never used that—I've got Windows 7, and that lets you rename multiple files by highlighting them all and then hitting F2; so I name the photos before I upload them.

Once you've copied the information, you can still make further tweaks—for example, adding more categories to selected files, or fiddling with the description of each file. For example, I'll often describe the first file like "Smith-Jones house, at the intersection of First and Euclid Streets in Sedgwick, Nebraska; seen from the". I'll copy that description to all of my photos of the SJ house, then put the direction for each photo in individually. Easier than typing the whole business over and over again. — Ammodramus (talk) 20:48, 20 October 2015 (UTC)

With my 10mb files, I have to do 5 or maybe 6 at the time, or it locks up and I have to start all over. At least that has been my experience. Bubba73 ^{You talkin' to me?} 02:52, 21 October 2015 (UTC)

Some of the things you said make it easier, but it still doesn't work right. I selected 26 photos to upload to the commons category "Downtown Waycross Historic District". It uploaded them all, but then it would not finish processing the last one. So I had to abandon that. I've had problems trying to do more than five or six at the time, so I started over and selected the first five and that went OK. Then I clicked on "add more files" and selected the next five. It processed all but one, and wouldn't finish processing the last one. This is the old problem with it not finishing processing more than about six files. So I had to abandon that and start over. I did them in five batches (5/5/5/5/6), which is tedious and time-consuming. Also, I didn't do any cropping or tweaking of the photos as I usually do. And they don't have good descriptions. Perhaps some industrious person will fix the descriptions, but I don't want to take the time... Bubba73 ^{You talkin' to me?} 05:54, 21 October 2015 (UTC)

I agree with Ammodramus. See the images in Commons:Category:East Second Street Historic District (not an NR-listed district, just locally created), most of which have similar descriptions — Front and DIRECTION [northern, western, etc.] side of the house located at NUMBER STREET [e.g. 612 S. Fess Avenue] in Bloomington, Indiana, United States. Built in YEAR, it is part of the locally-designated East Second Street Historic District. I took lots of these photos together and uploaded them in batches, but not with the batch uploader: I just prepared the description pages offline, opened a couple dozen browser tabs of Commons:Special:Upload, picked each image to upload in a different tab, pasted the description for each one into the appropriate page, and clicked "upload". This might be a better route with your larger images, since each tab is independent; if it doesn't process one, the rest are still good. Bubba, remember that Commons files can always be renamed after upload; the criteria are generally rather strict, limiting the renames to problems like typos, derogatory filenames, and non-descriptive filenames, but the first criterion is "At the original uploader's request". If you find it easiest, just upload them with numbers and then leave me a note saying "File 1 needs to be X, file 2 needs to be Y, etc." and I'll happily move them. Nyttend (talk) 02:09, 23 October 2015 (UTC)

I've had file renamed on Commons, but because of an error or a better or more accurate description. I think it would be easier to name them before uploading rather than after. Also, I think my problem with loading ~10 files of about 10MB each is due to a 100MB upload limit. I've selected the option "upload in chuncks" which does 5MB at the time for a total of up to 1GB, according to what I read (but not tried yet). Bubba73 ^{You talkin' to me?} 04:04, 23 October 2015 (UTC)

I spent 3 hours and 16 minutes in Ware County photographing (not counting a dinner break). I hit all seven HRHP sites (plus a couple more things) and took 288 photos. Over the next few days I uploaded about 81 of them. I think that took at least 12 hours. Bubba73 ^{You talkin' to me?} 04:48, 23 October 2015 (UTC)

I definitely know how that is; see my comments in the previous section regarding Indiana. I've made two trips recently, one to the Detroit area to expand my personal collection (a week ago yesterday) and another to finish off National Register of Historic Places listings in Richland County, Ohio (Tuesday), and even though there's no uploading needed for Detroit, I've still not yet completed the image processing for those ones. And I'm up earlier than normal today so I can get some sites in Columbus and points south (sites that Aesopposea illustrated long ago), as well as going to the Circleville Pumpkin Show for the first time since whenever it was that I took the picture that's in the article. Just digging myself deeper into the hole, making it longer and longer until I get the H. Lauter Company Complex image uploaded. Nyttend (talk) 10:50, 23 October 2015 (UTC)

In the process, I also created three or four stub articles to hold the photos and created commons cats and linked to them from the county NRHP article - quite time consuming. My hat is off to whoever recently created articles for all of N.C. (I forgot who did that.) Bubba73 ^{You talkin' to me?} 00:13, 24 October 2015 (UTC)

Question about a separate article

I visited Sapelo Island today, which has some NRHP sites. One is Sapelo Island Light. Sapelo Island Range Front Light, which is about 600 feet away, is discussed on the NRHP form. The latter article is very stubby. (1) Should it be combined with the former article? (2) If not, should it use the NRHP infobox? and (3) I'm going to upload two photos of it - should they be in the same category as the main lighthouse, or a new category? Bubba73 ^{You talkin' to me?} 23:27, 24 October 2015 (UTC)

My sense is that individual lighthouses from the 19th century are so few and far between that they are individually notable, so two articles are justified. Note, for example, the seperate articles for Delaware Breakwater Range Rear Light, paired with Delaware Breakwater Range Front Light, and Staten Island Light, paired with West Bank Light. If anyone's monitoring Wikipedia:WikiProject Lighthouses, they might have more informed opinion. If you keep two articles, the range light should get a "contributing property" infobox (set the parameter nrhp_type = cp), and I guess I'd keep photos in the same category, since the Range Light is part of the Sapelo Island Light listing, even if it has its own article. Andrew Jameson (talk) 11:00, 26 October 2015 (UTC)

OK, and old range lights are rarer than regular lighthouses. This is also believed to be the oldest surviving iron structure (of any kind) in Georgia. Thanks for telling me about the contributing property parameter - I didn't know about that. Bubba73 ^{You talkin' to me?} 14:48, 26 October 2015 (UTC)

I added the NRHP info box and fleshed out the stub some. Bubba73 ^{You talkin' to me?} 17:37, 26 October 2015 (UTC)

I've been making a lot of fixes to locations lately, some big, some small, and usually to make sure that the map shows the correct house. For many info boxes the location is in DMS format, as this is what elkman's tool emits. The county lists are in DEC format. Fixing the location in both the article and the county list (and the photo in the case where the upload wizard was used) is then much more difficult than need be. It would be easier if the article was in DEC format, and I've started doing that when I make a correction, for example, James S. Lakin House (diff). I'm not advocating a mass edit, just asking, does anybody (or anybot) care about this one way or the other? Unless | coord_format = dms is added, the location in the article now displays as DEC when it had been DMS. Does it matter? Generic1139 (talk) 18:58, 27 October 2015 (UTC)

When I've photographed a site, I get the coordinates from Google's satellite view, so they come in DEC format. I don't see any reason to change them to DMS. When I use one of the photos to illustrate an article, and find that the coordinates in the article need to be corrected or tweaked, I use DEC rather than DMS. However, I have no principled opposition to DMS: in those rare cases when the NRIS coordinates actually hit the right building, I leave them alone in DMS.

I almost never use the upload wizard from the county list. From Generic's comment, it sounds as though it automatically adds the site's NRIS coordinates to the uploaded file. If so, this is decidedly not a good thing. NRIS coordinates are often a mile or more from the right location, and even if they hit the property, they often hit a corner of the lot rather than the building of interest. I'd like to suggest that the people in charge of the wizard kill this. Better no coordinates than inaccurate ones. — Ammodramus (talk) 20:06, 27 October 2015 (UTC)

Gosh, it happens that I've been correcting a lot of coordinates yesterday and today! There are all kinds of problems. I find that many are 100 to 300 feet off - which can put it one or two blocks off in a city. Now if is something like a courthouse, that should be close enough to see it, but not if it is a smaller building or house. Around here, quite a few seem to be about 100 feet off (or 2") in the N/S direction, which may be a result of the change in maps years ago. Sometimes there is a typo. Sometimes longitude is only to the nearest 0.01 degree. Sometimes they seem to have just picked the nearest intersection. I had one that was a kilometer off for some reason. (The old vacuum-tube GPS units must not have been very accurate.) Also, most of the decimal degrees have repeating digits, indicating that they were converted from DMS to decimal degrees. I have seen at least one case of degrees plus decimal minutes. These things get transformed from the Easting & Northing into DMS and then often into decimal degrees. Anyhow, I prefer decimal degrees. I've been using ACME mapper (satellite view - and thamks to the person who told me about that) to get the decimal degrees to 0.00001 degree, which is about 1 meter. That is also more accurate than an integer number of seconds. Bubba73 ^{You talkin' to me?} 20:28, 27 October 2015 (UTC)

It's no better out in the country. I'm trying to improve our photo coverage of South Dakota, and a lot of the NRHP listings are barns or country schoolhouses or homesteads. I click on the coordinates, and the map puts me in the middle of a cornfield. Does this mean that the place has been torn down, or just that the coords are badly off? With the older nominations, the addresses are no better: we get things like "North of Highway 52" or "Rural Route 2". Location of these rural sites is often given by TRS1/4, and I've got maps of that for Nebraska, but can't seem to find corresponding maps for SD. I've occasionally found AR or wrong-addressed houses in small towns by sketching the house from the nom form photos, then roaming the streets until I find a house that matches. Think I'll probably wind up doing the same before I can kill off, say, Yankton County... Ammodramus (talk) 21:12, 27 October 2015 (UTC)

Ammodramus, check the ACME Mapper in topo-maps mode: you'll be able to see the section number, and if you go far enough along the border between two townships and the border between two ranges, you'll eventually come to the corner of the paper original map, where the township or range number for each one is illustrated. See [1] for a random example; the townships are marked a few pixels to the left of the center of the image. From there, it's easy to figure out the applicable PLSS location. Original question — I'm in favor of using DMS when possible, simply because it's the standard way of giving geocoordinates, unlike the decimals that are largely meant to make simple computers happy. Nyttend (talk) 00:14, 28 October 2015 (UTC)

Oh, and Bubba73, have you checked the comment about Florida in {{NRHPGoogleMapFootnote}}? Obviously you're a bit farther north, but that problem should explain what you're encountering; 100 or 200 feet is 1 or 2 seconds (a minute of latitude is a nautical mile, which is about 6,000 feet), so it's probably related to that as well as to the old-style maps. Nyttend (talk) 00:18, 28 October 2015 (UTC)

Thanks, Nyttend—that'll make SD a lot easier. Ammodramus (talk) 01:37, 28 October 2015 (UTC)

Responding to a couple of things, plus a comment: No, I haven't looked at the comment about Flordia, but I will. Second, "I click on the coordinates, and the map puts me in the middle of a cornfield" At least two of the ones I clicked on near here put them out in the ocean! (I fixed them). And in the ones I've checked, quite a few are off primarily in the N/S direction, although over half of them are off in a random direction. And a comment: a big part of the problem is that 20+ years ago, no one anticipated that people would be using personal GPS units to go to a precise location. Bubba73 ^{You talkin' to me?} 02:14, 28 October 2015 (UTC)

Good point. 20+ years ago, they really only needed coordinates in order to supply locations to CRM people and others who needed to know NR site locations for Section 106 checks prior to construction projects. Your ocean sites aren't the only ones; per Wikipedia:WikiProject National Register of Historic Places/NRIS information issues, the NRIS coords for Casselman Bridge in Maryland placed it on the equator midway between Ecuador and the Galapagos Islands. The equator I can understand (someone simply forgets to enter latitude data, so it defaults to zero), but the longitude was many degrees wrong, and the combination of the two mistakes was simply laughable. Nyttend (talk) 02:31, 28 October 2015 (UTC)

I prefer decimal degrees for a couple of reasons. (1) I use the ACME thing to get the lat/long, and it gives it in decimal degrees, (2) it is easier to put decimal degrees into my GPS. Also, the values given by ACME have five digits after the decimal, which is 27+ times more accurate than an integer number of minutes and seconds (although you can put in decimal seconds). And even four digits after the decimal point is 2.7+ times more accurate than integer minutes and seconds. Bubba73 ^{You talkin' to me?} 02:38, 28 October 2015 (UTC)

Thanks for all the input. The GPS in one of my cars accepts only DEC for coordinate input, the other accepts only DMS. I always include both now in my pre-trip note taking. Fortunately, the GeoHack app that comes up when you click any of the wikipedia or commons coordinate links gives both formats at the top. Since I haven't seen any technical reason to choose one over the other when editing the NRHP infobox, and since format conversion is an opportunity for error, I'm going to switch over and use DEC for the location in infoboxes I edit, since I've never seen anything else in a county list. I can then more easily propagate the change to list/article/photo edits. Generic1139 (talk) 05:35, 28 October 2015 (UTC)

Over the lawst 2 or 3 days, I've been going through the counties in Georgia of which I'm familiar and correcting coordinates. An hour or two I had a hard time getting the right location of the Pace house. (To make the process harder, my memory was telling me that it was on the west side of the street instead of the east side.) I went back to the NHRP form (2003 I think, so it should have the modern coordinates). Compared to the center of the house, the Easting was off 11 meters and the Northing was off 213 meters. Perhaps someone wrote 3526680 instead of 3526880.

Anyhow, most (by far) of the ones I've changed are in decimal degrees. Almost always only the first three digits after the decimal point are correct, and sometimes the third one is off by 1. So typically they are off maybe 0.0005 degree, which is about 180 yards. Bubba73 ^{You talkin' to me?} 05:55, 28 October 2015 (UTC)

Back in the day, we had to rely on just a topo map and our best estimates. Then we had the Internet and could use websites such as TopoZone, where we could hover the cursor over the item and get the coordinates. Now we have GPS. Einbierbitte (talk) 13:43, 28 October 2015 (UTC)

I've corrected probably a few dozen coordinates over the last 3 days. A few are way off (like over 800 feet) but they all seem to be at least 100-300 feet off of the actual building. As I said before, in the decimal ones, anything past the first three digits after the decimal point is garbage. Sometimes the third one is off by 1 or even 2. I'm suspecting a systematic error somewhere in the process of getting the coordinates.

For example, Ashantilly was added this year, so there should not be a problem of the maps using the old coordinate system. But it initially had coordinates of 31.3835, -81.4147. Using ACME mapper, I put it on the building and got 31.381165, -81.413051. So here even the third digit after the decimal was wrong. So, since they are all like this, I think there may be a bug in the calculation from easting/northing to lat/long. Bubba73 ^{You talkin' to me?} 14:52, 28 October 2015 (UTC)

Articles that are too sad to read

Is it just me, or are there some articles on Wikipedia that are so devastating, they'll bring you on the verge of tears? I often have a hard time dealing with the Anderson Street (NJT station) article, knowing the tragic outcome of that former landmark. ---------User:DanTD (talk) 21:18, 27 October 2015 (UTC)

A couple of weeks ago I read the article about the NRHP bill. It protects things from Federally-funded programs, but otherwise there doesn't seem to be any real protection, just a suggestion. Bubba73 ^{You talkin' to me?} 02:40, 28 October 2015 (UTC)

Want to really get depressed? Try Jobbers Canyon Historic District in Omaha, demolished except for one building to make way for an expanded corporate headquarters. It's the largest single loss of NRHP-listed property (although I would qualify that as the largest man-made loss; as the remnants of Hurricane Patricia pass over my region I can recall the losses inflicted by some of our recent storms, which further reminds me of how the Old Blenheim Bridge was swept irretrievably away by Hurricane Irene four years and change ago). Daniel Case (talk) 17:40, 28 October 2015 (UTC)

Not as bad, but I was working on List of Carnegie libraries in Georgia - some of these are on the NRHP. But five of them have been demolished. The Atlanta Main Branch looked like a beautiful building. Bubba73 ^{You talkin' to me?} 17:53, 28 October 2015 (UTC)

Fall 2015 NHL meeting

The NPS now has the agenda up for its next NHL meeting, scheduled for Nov. 16–17 at the Charles Sumner School in Washington. The following applications will be considered (Wyoming, our least populous state, seems to be a big winner in this round, with two new proposed NHLs, along with Connecticut).

Along with them are my comments about how well-prepared the associated articles are, if they exist. For most of them we are in good shape, but there are a couple where we'll have to start from scratch.

1. Ames Monument, National Register of Historic Places listings in Albany County, Wyoming. Henry Hobson Richardson-designed memorial to completion of transcontinental railroad near the spot (tracks themselves since relocated). We have an article with a picture. Executive summary and nomination form
2. Athenæum (Das Deutsche Haus), Indianapolis. Well-preserved 1907 Romanesque-Renaissance Revival building that, in addition to being a long-term social focal point for the city's German American community, was home to the first institution for training phys-ed teachers. Article and photo all set. Executive summary and nomination.
3. Gaukler Point, currently listed as the Edsel and Eleanor Ford House, Grosse Pointe Shores and St. Clair Shores, Macomb County, Michigan. Designation reflects the importance/excellence of the estate's landscaping by Jens Jensen. Article and photo already. If it's designated under the name used we might want to consider renaming the article as well Executive summary and nomination.
4. Hell Gap Paleoindian Site, near Guernsey, Goshen County, Wyoming. Only known Paleoindian site at which every Plains Indian cultural complex (save Clovis) has been found. Article exists, though without photo, despite site not presently being listed. Executive summary but no nomination yet; I suspect they are redacting specific location information out of it.
5. Man Mound, Greenfield, Sauk County, Wisconsin. Only anthropomorphic prehistoric monumental earthwork in North America (I wonder if the NPS has its eye on WHS status for this ultimately). Listed, and we have a photo but no article. Probably should fix that before it gets NHL status. Executive summary and nomination.
6. James Merrill House, Stonington, New London County, Connecticut. House of prominent writer who was among the first American writers to explore his homosexuality in print. Would be the third NHL to be designated for its importance in U.S. LGBT history after the Stonewall Inn and Henry Gerber House. Not listed yet but there is an article and photo, although it's grayscale so we should try to take a color one. Executive summary and nomination.
7. Mississippi State Capitol, Jackson, Hinds County, Mississippi. National significant example of Academic Classical Revival architecture by Theodore Link. Article and picture already (if there wasn't, something went seriously wrong a long time ago). Executive summary and nomination.
8. Norman Film Manufacturing Company, Jacksonville, Florida. Rare surviving example of a silent film studio. Not listed, no article or picture. Executive summary and nomination.
9. St. Bartholomew's Church and Community House, New York, New York. St. Bart's, with its Byzantine architecture not commonly associated with Episcopal Church buildings, is one of the few breaks in the high-rise landscape in that part of Midtown. As expected for a church so well-known (take a look at the movies and TV shows it's been used in) in our largest city, we have an article and plenty of photos. Executive summary and nomination.
10. The Steward's House, Foreign Mission School, Cornwall, Litchfield County, Connecticut. We have an article on the school, which was meant to train prospective missionaries of non-European background to proselytize amongst their own but closed down after a few years in the early 1820s. We don't have an article on the building, probably the best-preserved remaining structure (it's still a private residence) specifically. Nor do we have a photo, although this is not so far away from where I live as to give me the idea that it might make a nice excuse to drive to and photograph next month. Executive summary and nomination.
11. Zoar Historic District, Zoar, Tuscarawas County, Ohio. Well-preserved remaining community of German Separatists that is a popular visitor attraction in the area. Listed but we don't have an article; I suppose that although the HD does not cover the entire incorporated village this may be another instance where we have to consider whether the settlement and HD are identical enough to not need one. We do have pictures, though. Executive summary and nomination.

Daniel Case (talk) 18:35, 28 October 2015 (UTC)

We have color photographs of the James Merrill House, they're just not being used. Magic♪piano 19:05, 28 October 2015 (UTC)

Couple comments: (1) I would recommend against renaming the Edsel and Eleanor Ford House article as "Gaukler Point" - although the latter may be technically more correct, no one actually uses that name, and the "Edsel and Eleanor Ford House" is generally understood to include the surrounding estate grounds. I've attended several functions at the Edsel and Eleanor Ford House, but none have been in the house itself, just on the grounds or in another building. (2) The Man Mound is cool as hell - I've visited it, and didn't realize until now that it was on the Register, so it's fun to see it being considered as a NHL. Andrew Jameson (talk) 11:16, 29 October 2015 (UTC)

Very surprised that Zoar (rhymes with "lore", if I remember rightly) wasn't already an NHL; somehow I though they'd given it NHL status a few years ago. I've done a lot of Ohio writing (it has the third-largest number of start+ articles), but Tuscarawas County isn't a part of the state where I've done much of that writing; I tend to work mostly with individual buildings/structures/sites, and I've written very few district articles. We might want to look into sources discussing recent flood-protection issues; the nearby river dyke has been thought to be in danger of flooding, if I remember rightly. Nyttend (talk) 18:41, 29 October 2015 (UTC)

Norman Film Manufacturing Company has an article and some pictures under Norman Studios. If got added to the NRHP on December 29, 2014 as Norman Film Studios. --‖ Ebyabe ^talk - Union of Opposites ‖ 05:19, 3 November 2015 (UTC)

I found some time today to write it, so Man Mound has an article now. TheCatalyst31 ^{Reaction•Creation} 22:36, 15 November 2015 (UTC)

San Bernardino Pictures

This weekend I went out and got a bunch of pictures in San Bernardino County. Among the surprises was the Ontario State Bank Block. It was demolished, probably when Euclid Avenue was lowered under the railroad tracks. Main Street (the corner where the bank stood) no longer goes through

Now an empty lot

Einbierbitte (talk) 04:53, 2 November 2015 (UTC)

EDIT - so what do we use for pictures? An empty lot? Einbierbitte (talk) 14:53, 3 November 2015 (UTC)

That has been done in some cases. It helps the reader and other editors understand why they can't locate the building when they visit. Jonathunder (talk) 16:31, 3 November 2015 (UTC)

Actualy, this page says it was demolished. I will add it to the article. The illustration of the lot will be in order there.--Ymblanter (talk) 00:19, 4 November 2015 (UTC)

While I'm all in favor of adding photos of empty lots to the Commons where an NRHP building has been demolished, Wikipedia is not a travel guide: we should focus on illustrating the significance of the listing, not helping people find the place or understand why they can't. Meanwhile, the nomination photos for this building are in the public domain, and I have uploaded them. — Ipoellet (talk) 03:15, 4 November 2015 (UTC)

After reading the paragraph on travel guides you linked to, I don't think this would be a "travel guide" problem. Bubba73 ^{You talkin' to me?} 04:22, 4 November 2015 (UTC)

"Not a travel guide" is meant to prevent us from adding things like directions to a site, hours, admission fees, etc. It's not at all meant to illustrate the current location of a site. Given the absence of proof for the claimed PD status for these nomination images, I've nominated all of them for deletion. Nyttend (talk) 04:28, 7 November 2015 (UTC)

Coordinate errors

Over the past couple of weeks, I've corrected probably a few dozen coordinates of NRHP sites by finding the building using ACME Mapper and reading off the actual coordinates. (I'm assuming that ACME Mapper is accurate - it agrees precisely with Google Maps.) Anyhow, when the coordinates are given in decimal degrees, almost always everything after the third digit after the decimal point is wrong, and when they are given in D/M/S, they are no better.

Here is an example, from National Register of Historic Places listings in Effingham County, Georgia, the jail. The NRHP form is from 2006, so it should have recent (accurate) data. It gives the UTM of zone 17, easting 470266 Northing 3581760. It looks like it might be accurate to within a meter.

Using this to convert to degrees, it gives 32.372413°N 81.316078°W / 32.372413; -81.316078. The page lists 32.374167N, 81.316111W. Notice that everything past the third digit after the decimal point is wrong, which is typical in the ones I've examined. In this case, the north coordinate is off by 0.00175+, which seems to be larger than typical.

So the first point is that the coordinates given in most places do not agree with the conversion from the NRHP forms after the third digit after the decimal point, which is enough to throw them off a block or two. I suspect that when these were converted from the UTM on the NRHP form, the conversion wasn't done correctly. (I don't know who did these or when.) Does anyone know details about how these conversions were done?

The second point is that the NRHP UTMs are quite a bit off, assuming that the conversion website above and ACME Mapper are accurate. The coordinates given on the NRHP form, converted to degrees, puts the jail out in the woods, whereas I think the center of the building is actually at 32.37413°N 81.31570°W / 32.37413; -81.31570. Using the conversion website, that gives UTM zone 17, easting 470302 northing 3581950 - a difference of 36 meters E/W and 190 meters N/S. (The courthouse is off even more, but its NRHP form seems to be rounded to 100 meters.) Bubba73 ^{You talkin' to me?} 23:36, 2 November 2015 (UTC)

Elkman's infobox generator uses a good algorithm, so I wish I had access to it, because the online UTM-to-DMS converters that I've found are routinely badly wrong. I've not yet found a converter that I trust at all solidly. Nyttend (talk) 03:32, 3 November 2015 (UTC)

I suspect that there was a significant error in the program used to convert these in the first place. I'm trying to think of a way to check the website I'm using (I think it is from some university so they should know what they are doing.) Bubba73 ^{You talkin' to me?} 03:38, 3 November 2015 (UTC)

I checked the website I use for conversions with this one from NGS/NOAA, and it agreed. Bubba73 ^{You talkin' to me?} 04:13, 3 November 2015 (UTC)

One degree of latitude is 111.2km, so .001 of a degree is ~111 meters. .0001 is 11 meters. That is close enough to get to within a house or two, north and south, in an urban area. Close enough to pick out a farmhouse. .00001 is 1 meter, and .000001 is just silly. One degree of longitude is harder to quantify as is depends on your latitude as well. It is ~111km at the equator, but only ~85km at 40 degrees latitude. So, in Pennsylvania, .0001 longitude is 8.5 meters. In DMS, a minute of latitude is ~1.85km, a second is ~31 meters, a tenth of a second is ~3 meters.

Sources of error:

• original determination from topo maps. These were done by drawing two lines from the up/down, left/right sides of the map and interpolating what the northing and easting values are - accuracy depends on how hard the map reader worked at it. Meter resolution is not to be expected, this is why surveyors go out into the field to measure important things.
• Datum - the corrections for the shape of the earth - for the original map, and the digital map you are using. Differences of 10s to 100s of meters are possible.
• Calculation - depending on the original source of the location, by the time it gets to the county list, it might have been moved across different datums, and from DEC to DMS, more than once, with loss of data on each conversion.
• What is being reported on the original nomination. Sometimes, the center of the plot of land is being reported, not the house. The bigger the plot, the more likely this is true. Sometimes, for a large house, the location doesn't point at the middle. Some editors (and sometimes me) tend to try to reduce false precision by shortening the number of digits in the lat/lon of point and click map-determined locations, accepting a shorter number that falls within the boundary of the building.
• What the editor that created the county list used. Sometimes, there is no location in the registration form, or the regestration form is not avaialble, but a street address is. If the editor enters that street address into Google, Bing, or some other online mapper, the result can be wrong - wrong side of the street, wrong side of town. Address schemes change over time.
• And, sometimes, the data is just wrong. Transcription errors, misreading of the original easting/northing registration marks on the map, whatever.

Finally, as has been discussed here, the upload wizard (what you get when you click on "upload image", takes the location from the county list, and copies it into the location information when the image is created in commons. That propagates errors into commons.

So, what am I getting on about? Don't talk about decimal places out to the 4th digit and beyond as "wrong" - the system for generating them, for the reasons above, aren't ever going to match ground truth. What we may need is a "verified" tag - meaning that someone has gone to the trouble of checking to see if the location in the county list is correct - that photos, maps, streetview, whatever is available, have been checked against the location. And, for non-HD listings, the tag says that the location points at (one of the) primary building(s) in the listing, and not the center of a 4000 acre plantation. And, at some point, run a bot that extracts the location in the county list and the location in the article, checks to see if there are within some fuzz value, and makes a list of those that aren't. Generic1139 (talk) 17:00, 3 November 2015 (UTC)

I'm still learning about the technical aspects of the mapping system. But the coordinates are off by a lot more than 0.0001 degrees (~11 meters). They are off several times that amount. I just corrected several in Nassau county, Florida, and they were all off at least a block, and as much as four blocks off (that one was probably a clerical error somewhere). That makes it hard to find a house. Add to that the fact that many houses do not display their street number, so you often can't find them by their address. And many of the NRHP forms are not available, to give you a picture of what to look for. Anyhow, I believe that all of the coordinates listed in the county NRHP were done incorrectly somewhere - the coordinates in the NRHP forms don't correspond to what the county listings have (and usually that is the same one used in the article, if the site has one). Bubba73 ^{You talkin' to me?} 17:43, 3 November 2015 (UTC)

In Genesee County, Michigan, the coordinates for Atlas Grange Hall were actually the Barn at 4277 Irish Road (which also had the same, correct coordinates). I've always just corrected these on my own; I can't find any actual sources for the coordinates. I understand errors occurring from calculations, conversions, pre-GPS technology, etc, but what I don't understand is where all the coordinates in these tables are coming from in the first place. The pages linked in the tables are useless for Michigan, and the NPS website isn't very user friendly, anyway. kennethaw88 • talk 02:51, 4 November 2015 (UTC)

Ipoellet got Swayze Apartments spot on, I suspect that one came from the street address with a tweak from street view to get the right building, maybe they'll let us know in this case. Michigan had, at one time, a web site with information, the older ones could have come from that. The addresses are usually given in the weekly action listings, like this one. And I should have added "cut and past errors" to the list above. Generic1139 (talk) 15:49, 4 November 2015 (UTC)

That is precisely how I did the coordinates for Swayze Apartments: The Weekly List gave me an address, which gave me the location of the building. Google Earth gave me the lat/long coordinates necessary to express the same location in a different way. And yes, I probably used Street View to double-check that the Google address search correctly took me to the right place - their database isn't perfect.— Ipoellet (talk) 22:03, 5 November 2015 (UTC)

I'm aware also of a handful of errors in North Carolina listings which I've been meaning to correct, but haven't gotten around to. Sometime soon, perhaps... --Ser Amantio di Nicolao^{Che dicono a Signa?}_{Lo dicono a Signa.} 02:35, 5 November 2015 (UTC)

The errors are significantly more than 0.0001 degree - I estimate 0.0003 on the average. And the errors are in both coordinates, making the typical error on the order of 150 feet. They are just about always at least 100 feet off and often 300 feet.

The errors make it hard to find the correct house. In one city I had the coordinates and the street address of a house. I got there and couldn't tell which house was the right one. I thought "it must be one of these three" and photographed all of them. When I got back I looked up the NRHP data, and it wasn't any of them.

This is a particularly extreme example (I haven't fixed it yet): the courthouse in Effingham County. The actual coordinates are 32.37416°N 81.31487°W / 32.37416; -81.31487. The county listing gives 32.37548°N 81.31785°W / 32.37548; -81.31785 - 0.0013 degrees off N/S and 0.003 degrees off E/W - roughly 1,100 feet off. Its article gives 32.37722°N 81.31778°W / 32.37722; -81.31778, 0.003 off N/S and 0.0029 off E/W - roughly 1,500 feet off. (The jail is way off too.)

The coordinates given in Taylor Hall (Hawkinsville, Georgia) from the coordinates given (which are 4" north of the NRHP form, which aren't right either). When I was there, I couldn't find it by address (just "Kibbe St" at that time) or the coordinates. Bubba73 ^{You talkin' to me?} 02:42, 5 November 2015 (UTC)

Facts in Wikipedia are supposed to have references, but we hardly ever know who calculated our coordinates. When I correct by Google Earth (which in my experience holds steady for New York City, and often jitters year by year by tens of meters for Pittsburgh) I do not systematically say so. Sometimes the coords of pictures are from EXIF, which minimizes transcription errors but can create large GPS errors. Sometimes the coords for a picture are merely the center of town, even for a suburban object. Occasionally it's the location of the museum that holds the picture of a place far across the sea. Yes, we need some sort of standard or system for reporting the sources of coords. Jim.henderson (talk) 03:23, 5 November 2015 (UTC)

Since we're sharing...I was in Williamston, North Carolina a few months back, and ended up photographing all of the NRHP sites in town to add to Wikipedia. To do so I looked up images online and then matched them to the Google streetscape. What I discovered was that a handful of the coordinates are off. Three examples: Asa Biggs House and Site (it's listed as being about a block away from its actual site); Martin County Courthouse (waaaaaaay off); and Sunny Side Inn (nearly accurate, and I'd let it slide were it not for the fact that the business is actually on Google Maps as an identified landmark.) Lord only knows how many other errors there are in Martin County alone, let alone the rest of the state. I'm an outsider there; I wouldn't know where to begin beyond fixing the ones I've listed. --Ser Amantio di Nicolao^{Che dicono a Signa?}_{Lo dicono a Signa.} 06:35, 5 November 2015 (UTC)

It has been the loose recommendation of the NRHP project that locations should be fixed, or adjusted, as needed to point at the object that the article is about, in both the county lists and the articles. But, as Jim.henderson (most recently) reminds us, facts in Wikipedia are supposed to have references. Sadly, one of important aspect of a historic place is "where is that place", and due to all the reasons discussed above, that piece of data is often imprecise, and sometimes flat wrong. Since the project strives to include images for the articles, we're often in the business of finding a place with enough accuracy to correctly identify the object, and with sufficient precision so that it can be picked out of a crowd of nearby similar items. We usually know, therefore, when the original data is wrong. In the normal case, a reliable source is needed for facts, but in our case, we know that the "fact" is incorrect. The whole point of reliable sources is that no one cares about the editor's "facts", just citable source facts. I think the main underpinning of our "go ahead and fix it" philosophy is that the truth of the "where" is ground truth - there it is on a satellite map, or in the case of a demolished property, there it isn't.

We make other types of location changes as well. A 1970 nomination form might give the location as Main Street, but the address in 2015 is Martin Luther King Jr. Boulevard. We sometimes change the address in the infobox to the new address, we sometimes note the original address, but I don't think we ever cite the local city ordinance that made the name change.

The point is, do we accept google/bing/others maps as a sufficient reliable source (implicit or explicit) and not worry about citing the source for the location? Or do we come up with some sort of tag or parameter in NRHP Row and the NRHP infobox that marks a correction or original placement done by the editor, and cites the method used, such as:

• Street Address
• Satellite Imagery
• Street View Imagery
• Original nomination drawings

Going back over 90,000 or so locations is not likely, but moving forward it would be possible. Generic1139 (talk) 16:41, 5 November 2015 (UTC)

Facts may require citation, but trivially* verifiable ones don't (or shouldn't; if they do, then WP:V is IMHO broken). The location of a building or structure, especially if it is visible from a publically accessible location, is trivially verifiable (by going there). That said, I agree that it would be useful to have a means (in both articles and lists) to inform how a particular set of coordinates were provided, including how, when, from what source, and by whom. This information does not necessarily need to be visible in a normal rendering of the table or infobox.

(*By "trivially", I mean not requiring any unusual tools, knowledge, or access to specialized or restricted documentation. This listing's location is somewhat expensive and logistically complicated to verify in person, but the average person with time and means can in principle do so, and is IMHO trivially verifiable.) Magic♪piano 17:39, 5 November 2015 (UTC)

Just a quick note - a few of the NRHP forms that I've examined and compared give the coordinates in D/M/S instead of UTM easting and northing. Except for one that was precisely 1' off (probably a clerical error) the other forms that give D/M/S seem accurate enough. That, plus what I've discussed above, makes me believe that there was a serious error in converting the UTM into lat&long. I don't know how the NRHP forms were converted from UTM to lat&long, but if they are in a database, a better conversion can be done. As far as verifiability, a proper conversion from the NRHP UTM data would be verifiable and a lot more accurate than what we have now. Bubba73 ^{You talkin' to me?} 18:02, 5 November 2015 (UTC)

@Bubba73: please provide some examples where you suspect a flawed UTM conversion, let me see if there is any common factor or other reason to believe there is a systemic UTM to DMS conversion problem. @Magicpiano: I agree that the source of the coordinates need not be visible (without clicking something). Generic1139 (talk) 18:54, 5 November 2015 (UTC)

@Generic1139: I feel like you're getting lost in the details. The coordinates and address of a listing are not a "fact" subject to WP:V (at least not the single-point coordinates in lists and infoboxes). The key "fact" is the location of the listing - the address and coordinates are just two of several ways of expressing that location to our readers. Once the location is established with a verifiable source (nearly always the listing's nomination form, AR aside), then it is entirely within the editor's remit to find different ways to express the location that best suit the context and best convey the location fact to readers, just the same way that an editor might reword a fact in a prose paragraph. Starting to offer up citations for every step of adjusting coordinates, updating postal addresses, or re-expressing an address as coordinates I feel is a form of TMI that makes the encyclopedia less useful to readers. If an editor cites a source that gives the location, then takes care to make sure any coordinates or addresses are a valid way of expressing/conveying that location, then that ought to be sufficient.— Ipoellet (talk) 22:48, 5 November 2015 (UTC)

@Generic1139: OK, I will by going through my recent changes, but it will take some time. It would be better if I'd kept a record of these. I wrote the UTMs and the actual coordinated down on scratch paper but I didn't write down what object they went to. Bubba73 ^{You talkin' to me?} 23:26, 5 November 2015 (UTC)

Examples

• National Register of Historic Places listings in Wilkinson County, Georgia, the Elam-camp house. The 1982 form says: zone 17, easting:281380 northing: 3640560. This converts to 32.881396°N 83.336899°W / 32.881396; -83.336899. The county article had 32.883056°N 83.336944°W / 32.883056; -83.336944. The actual location (from ACME mapper): 32.88333°N 83.33679°W / 32.88333; -83.33679. The difference N/S is 0.00166 degrees, about 600 feet.
• National Register of Historic Places listings in Effingham County, Georgia - the jail. The 2006 form says: zone 17, easting 470266 Northing 3581760. That is 32.372413°N 81.316078°W / 32.372413; -81.316078, the article gives: 32.374167°N 81.316111°W / 32.374167; -81.316111, off by off by 0.00175 degrees N/S (over 600 feet).

• National Register of Historic Places listings in Effingham County, Georgia - Courthouse. the 1980 form: Zone 17, E470100 N3582100, which is 32.375475°N 81.317854°W / 32.375475; -81.317854. They look like they have been rounded to the nearest 100 meters, but they shouldn't be more than 150 feet off. The article had: 32.376667°N 81.317778°W / 32.376667; -81.317778 - a difference of 0.00119 N/S - over 400 feet off.

• National Register of Historic Places listings in Wheeler County, Georgia, the courthouse. The 1980 form: zone 17, E: 331900, N: 3558080. This is 32.146627°N 82.782394°W / 32.146627; -82.782394. The page had: 32.147778°N 82.7825°W / 32.147778; -82.7825 - 0.00115 degree off N/S, over 400 feet. (Actual: 32.14813°N 82.78230°W / 32.14813; -82.78230.)

I don't know what the cause of the errors is. If they used the old datum, the N/S error should be about 100 feet, but these are several hundred feet off in N/S, and the UTM on the forms is south of what it should be (by several hundred feet). But by 2006 (the second example), they should be using good maps. In these cases, the old lat/long coordinates in these articles are more accurate than the UTM from the forms, so the problem may not be in the conversion from UTM to lat/long, as I thought. But I don't understand where the error is. Bubba73 ^{You talkin' to me?} 00:53, 6 November 2015 (UTC)

Analysis of the first two

Here is what I found on the first example, Elam-camp. First, your conversion of E/N 281380, 3640560 to DEC appears to be incorrect, based on the numbers, I think your conversion may have assumed that the input and output were were NAD83. I get 32.881397,-83.336898 when I assume that, a difference of 0.000001, -0.000001 from your numbers. The correct conversion is from NAD27 to WGS84, resulting in 32.883259,-83.336778, a difference of .001863, 0.000121, or 207 meters.

The distance between the NPS E/N (converted NAD27 to WGS84) and your satellite location is 8 meters, which is not bad. I was using this to do the conversions.

But, why does the county list say 32.883056, -83.336944?

The original version of the page gave the location as 32 52 59 N, 83 20 13 W. That is rounded but from what? If you assume that the E/N was NAD27 and the output was NAD27, then you get 32 52 59.21945 N 083 20 12.81303, which, rounded is 32 52 59 n, 80 20 13 w. A NAD27 to WGS84 conversion gives a different answer. I don't know what the orignal assumption was, and am trying to find out, but, 32 52 59 n, 80 20 13 was what was in the original county list. A bot later converted that to DEC, resulting in what is now in the county list, 32.883056, 83.336944.

The maximum rounding error, N/S would be .5 seconds, or about 15 meters.

Summary for Elam Camp: original NPS data, when using the correct datum conversions, is 8 meters off of what you determined. The county list suffers from rounding to the nearest second, and also possibly from using an incorrect datum, resulting in an error of 184 meters, most of it from the possible datum issue 27 meters, some rounding, and some datum.

I'll look at the other two tomorrow. Generic1139 (talk) 06:44, 6 November 2015 (UTC)

Wheeler County Courthouse, similar story as Elam-camp

You have a typo in your example, you gave N: 3558080, the form says 3558030, when I use your data and NAD83 to NAD83, I get your number and a 158 meter error. With E/N converted NAD27 to WGS84 I get 32.148050,-82.782242, 10 meters from your satellite location.

DMS for the converted E/N is is 32 08 52.98N 82 46 56.0712W, rounded is 32 08 53N 82 46 56W. The county list had 32 8 52 N 82 46 57 W, again showing (I think) a NAD27 to NAD27 conversion, and rounding, resulting in a 38 meter error.

I think the larger errors you saw with these two examples are your incorrect conversion of the E/N data. Try again with NAD27 to WGS84 and see if you replicate my results. I see there are a total of 4 example, I'll also look at the other two. Generic1139 (talk) 07:44, 6 November 2015 (UTC)

Analysis of the second two

Effingham County, Georgia - the jail

Difference between E/N and orignal county list is 24.6 meters.

Difference between E/N and the Bubba73 update to the county list (32.372413,-81.316078) is 211 meters, but google doesn't show anything at that location - I think you updated the county list with an E/N converted with the wrong datum.

Difference between E/N and where I think the building is (32.374416, -81.316035) is 16.8 meters.

Effingham County, Georgia - Courthouse

Here is one that is further away than it should be.

Difference between E/N and the original county list is 18.7 meters

Difference between E/N and where it really is: 434 meters. Looking at the 1971 topo, the E/N is just wrong. The listing is from 1980, but the E/N could have come from a much older map. Anyway, there is nothing in the ingestion/conversion that would have helped here, the nomination form was wrong. The new GIS data from NPS, however, show a location for this building that is only off from satellite by 17 meters.

Final word from generic1139 on these examples

• Looking at these four, not a large sample for sure, and ignoring the method Bubba73 used to convert E/N to DEC in the examples, the Focus data might have an issue sometimes with datums used for the conversion. That issue will cause larger errors in some parts of the country. Rounding from E/N to DMS to DEC also causes a small error. These errors (15-30m) are large enough to mis-locate buildings in an urban setting, but won't be that far off.
• The source of larger errors is simple bad data in the first place.
• Our policy of adjusting the locations as we find problems is fine.
• Photographers should check the available data before setting out on a trip, locations from the data base, at best, are usually off by at least 20 meters or so, and there are outright errors of varying significant distances.
• NPS is now putting out better GIS data. We might be able to decrease frustration in the photographer ranks by finding a way to use it.

Fini Generic1139 (talk) 23:10, 6 November 2015 (UTC)

I've looked through this quickly - I need to go back and digest it (I was out photographing Effingham County yesterday - I corrected the coordinates of the courthouse and jail). I am using the 1984 datam for conversion, instead of the 1927 datam. But why would forms done in the last 10 years use the 1927 maps instead of the 1984 maps (e.g. the Effingham County Jail)? I don't know exactly how much the difference is in this error, but a graph I saw indicated that the difference should be a lot less than what I am seeing. And thank you for that link to the conversion. Bubba73 ^{You talkin' to me?} 16:17, 7 November 2015 (UTC)

Using that conversion website, I took a place near me and converted to 1984 coordinates and 1927 coordinates. The difference was 18 meters in easting and 209.6 meters in northing - a lot more than I had thought. The page that someone referred me to said that locations in Florida differ by about 100 feet, and a graph I saw showed Georgia deviations as less than Florida's. Bubba73 ^{You talkin' to me?} 16:28, 7 November 2015 (UTC)

Category:Historic Districts on the National Register of Historic Places by state

I propose to continue populating Category:Historic Districts on the national Register of Historic Places by state which I started as part of the new Category:Historic Districts on the National Register of Historic Places that I created. I created these categories to ensure that NRHP historic districts are in a category named as such and are properly placed within the Category:National Register of Historic Places category tree. This matches the high level category structure within Category:National Register of Historic Places where the category names include some form of National Register of Historic Places. Previous to my work, these articles were in Category:Historic districts in the United States by state subcats which included both NRHP historic districts and non-NRHP historic districts. My category scheme will clarity what is what at the category and name level: NHRP districts go into the Category:Historic Districts on the National Register of Historic Places in foostate. Category:Historic Districts on the National Register of Historic Places in foostate is a subcat of Category:Historic Districts in foostate since that is what they logically are: a part of the whole. Districts that are only non-NRHP districts go directly into Category:Historic Districts in foostate. Districts that are both NHRP and non-NHRP go both directly into Category:Historic Districts in foostate and its subcat Category:Historic Districts on the National Register of Historic Places in foostate--as allowed for in the WP categorization rules. Category:Historic districts on the National Register of Historic Places in Illinois illustrate part of this structure. Of course, categories are based on article text. This plan can do nothing about the general lack of content in article regarding non-NHRP districts. One editor objected to my work; so far, no one else has agreed with his objections. Your comments on my continuing to finish this category tree setup I just described. Thanks Hmains (talk) 06:48, 6 November 2015 (UTC)

Unlike Commons, I don't have much to do with categorization at WP, so I can't really comment on the proposal generally. However, it seems to me that the capitalization should be consistent: "Historic districts on the National Register of Historic Places by state" rather than "Historic Districts on the national Register of Historic Places by State." Lowercase the "d" in "districts", and uppercase the "N" in "National". — Ammodramus (talk) 11:48, 6 November 2015 (UTC)

Very true. I saw these last night as I was writing up my statement here, but did not want to confuse things at this point. I will change the capitalization before I add any more state level categories. Thanks Hmains (talk) 17:51, 6 November 2015 (UTC)

As I noted previously, virtually all of our HD articles are NR-listed. Moving the NR-listed ones into their own categories would leave each "HDs in PLACE" category with essentially no contents except the NR-listed HDs category for the same place. This proposal would add an additional layer of category complexity without providing any significant benefits. Finally, note that Hmains cannot be trusted to obey policy: he began this project despite opposition and without consensus in its favor, using the AWB software, despite policy's clear requirement not to use this software to make major/bold changes of this sort without solid consensus. Nyttend (talk) 04:13, 7 November 2015 (UTC)

Note that Nyttend's assertion that I did anything wrong in using AWB were questioned if not rejected by the editors/admins at Wikipedia:Administrators' noticeboard/Incidents#Hmains and AWB. Hmains (talk) 18:41, 8 November 2015 (UTC)

I do not have a strong opinion on the matter, but I think at this point the question can only be resolved via CfD.--Ymblanter (talk) 16:22, 10 November 2015 (UTC)

• Added information. Prior discussion here: Wikipedia_talk:WikiProject_National_Register_of_Historic_Places/Archive_62#Issue_with_categories_under_Category:Historic_districts_in_the_United_States_by_state. Other prior discussion: Wikipedia:Administrators' noticeboard/Incidents#Hmains and AWB. Hmains (talk) 06:37, 7 November 2015 (UTC)

I have been sorting NRHP articles into type categories (mostly west of the Mississippi so far) and I vigorously, vigorously support Hmains' proposal. This resolves an awkwardness that NRHP HD articles are floating at the top level of state categorization (Category:National Register of Historic Places in Wisconsin is a good illustration) whereas buildings/structures/objects can be finely sorted into categories by function (Category:Hospital buildings on the NRHP in Foostate, Parks on the NRHP in Foostate, etc). That there will be few non-NRHP historic districts to populate the higher-level Category:Historic districts in Foostate is consistent with the minimal content in categories like Category:Farms in Foostate or Category:Archaeological sites in Foostate vs the daughter categories Category:Farms on the National Register of Historic Places in Foostate and Category:Archaeological sites on the National Register of Historic Places in Foostate. etc. Plus the few non-NRHP HDs that do have articles will not be falsely categorized under the NRHP umbrella. That seems to me a far greater logical issue than some handwringing about underpopulated parent categories. -McGhiever (talk) 04:59, 10 November 2015 (UTC)

I saw the open discussion on WP:AN/I regarding Hmains' use of AWB. There didn't seem to be any appetite to sanction him on the basis of the discussion Nyttend cited, and I suggested he open a new thread here, because I thought that the situation might get more tense if it drifted off unresolved, and building consensus here could defuse that. I think this is a better place than CfD to hold the discussion, as it's more likely to attract the attention of knowledgeable editors with relevant experience.

@Nyttend and Hmains: Can you clarify to make sure I understand your respective positions correctly? As I understand it, for a historic district in county, state, Nyttend's scheme would have it classified under "National Register of Historic Places in county, state" and "Historic districts in state". Hmains' scheme would have it classified under "National Register of Historic Places in county, state" and "Historic districts on the National Register of Historic Places in state", which would be a subcategory of "Historic districts in state".

I admit I'm having a hard time rationalizing any strong preference for either. I think we can all agree with Nyttend that the size of the parent and child categories under Hmains' system would be lopsided, but I'm not sure that's so bad a thing we have to avoid it. But then, why subdivide "Historic districts in state" at all? Well, maybe the answer is that it's a good idea to subdivide "National Register of Historic Places in state", as McGhiever observed, and historic districts seem like natural children when dividing that category. So I guess I'm weakly in favor of the separate districts-on-NRHP categories, but I'm pretty open to persuasion. Choess (talk) 04:25, 11 November 2015 (UTC)

Is there any appetite for sanctioning him for the lack of consensus? AWB policy demands consensus in favor of any non-trivial action, not the lack of consensus against it. We'll still have plenty of articles on covered bridges, banks, armories, etc. left over after we move the NR-listed ones to assorted useless topical subcategories, but not the HDs. The current setup doesn't cause non-NR HDs and their CPs to be put into the NR tree; it's careless editing that does. What's more, having NR HD categories runs the risk of confusing the non-careless editors when NR-listed places are included in non-NR-listed HDs. Consider the John D. Haynes House of Fort Wayne, Indiana, listed on the NR as a building and comprising its own local HD. Given the existence of an NR HDs category for Indiana, it looks like a categorisation mistake, a situation that needs to be fixed by moving it from (HDs in IN) to (NR HDs in IN), which it isn't. This would be an error, and while it's not a problem we have with the traditional setup, it's a problem we have if we're having NR HD categories, a problem that's unique to HDs and not a factor with farms, archaeological sites, covered bridges, etc. Nyttend (talk) 05:16, 11 November 2015 (UTC)

PS, why must the NR HDs be floating at the top level? In some states, people have subdivided sites by county (e.g. Category:National Register of Historic Places in Allen County, Indiana), so NR listings of all sorts are moved into those categories when they exist; couldn't you just create some such categories? Nyttend (talk) 05:17, 11 November 2015 (UTC)

I strongly support Hmains' proposal.

• Issue #1: Everything under NRHP categories should be on the NRHP, so the "HDs in Foostate" categories should be removed from NRHP categories.
• Issue #2: Everything listed on the NRHP should be in an NRHP category for its location. This requires either separate "HDs on the NRHP in Foostate" categories, or having the NRHP HDs directly under "NRHP in Foostate" categories. I prefer the former. (Of course, where we say "Foostate", it could be one of the US territories, not just the actual states.)

It would be OK with me if HDs that are both NRHP-listed and listed elsewhere were in both "HDs in Foostate" and "HDs on the NRHP in Foostate" categories, as long as there is a notice in each category such as you get by using the template {{Non-diffusing subcategory}}. --Auntof6 (talk) 05:20, 11 November 2015 (UTC)

• I was asked somewhere above to further illustrate my work here. Category:Historic districts on the National Register of Historic Places in California is a good example. All Historic districts on NRHP in California are included here. The category name is no longer confusing: any reader can readily see what is contained here. Of course, this category is a subset of its now parent category: Category:Historic districts in California. Using this latter category allows the user to readily see which historic districts in California are not NRHP. This could not easily be done with the previous set up in which both NRHP and non-NRHP districts were simply jumbled together within Category:Historic districts in California. Any historic districts that are both NRHP and non-NRHPs go directly both in the subcategory Category:Historic districts on the National Register of Historic Places in California and the parent Category:Historic districts in California. This duel child/parent categorization is allowed within categorization rules. This change of mine puts Historic Districts into the same category structure as found with other NRHP categories. Example: Category:Buildings and structures on the National Register of Historic Places in Alabama within Category:Buildings and structures in Alabama and every other state in Category:Buildings and structures on the National Register of Historic Places by state. NRHP Historic districts category structure was the outlier here; now it will match up and be uniform with the rest of the NRHP structure. Hmains (talk) 02:00, 12 November 2015 (UTC)

• I conclude from the above comments that the consensus is that I continue populating Category:Historic Districts on the national Register of Historic Places by state as I described above in detail. That being so, I will re-start my work here in a few days. Thanks for all the supportive comments. Hmains (talk) 00:08, 16 November 2015 (UTC)

Focus KMZ files

NPS has a page that provides google earth .kmz files for NRHP listings up to April 2014. The page says "most of the building locations have been geocoded through a project undertaken with TeleAtlas which generally provided the most accurate locations (these have a "1" in the geocode match field). Locations that could not be geocoded, were plotted from the existing points in the database." I converted these to spreadsheets and looked at the samples from Bubba73 in the discussion on Coordinate errors; in both of the samples I looked at, the spreadsheet gave data that was better that the converted, rounded data in our county list. I'm not sure at this point how we'd easily use this new data to improve our lists. We could walk each NRHP Row, see if the location matches the database elkman used, and if it does, then use the new data if it is different (or build a review page). That would avoid overwriting our tweaked data. While we were bot-ing around, we could compare county list locations with the locations in the matching article and report differences > x feet somewhere. And, at least some of our "coordinates missing" lists have coordinates in the new data. Generic1139 (talk) 08:48, 6 November 2015 (UTC)

Emporia, VA NRHP list is complete

I'd just like the rest of you to know that I've added images to all NRHP sites in Emporia, Virginia. A few sites could use some additional images, some of which I intend to add, but at least none of the sites are empty now. If nobody has been there, there's a long row of historical markers on Main Street in that city, and one interesting house across the street from the masonic lodge. ---------User:DanTD (talk) 02:15, 9 November 2015 (UTC)

Best available satellite photos

Are the satellite photos on Google Maps and ACME Mapper the best ones available? (I assume so.) Sometimes I try to identify a house by its footprint, chimneys, etc, but sometimes the images aren't good enough. Bubba73 ^{You talkin' to me?} 21:35, 12 November 2015 (UTC)

I always shop around if there is a detail I need to see. Google and Bing sometimes have different overhead shots. Google street view and Bing bird's eye will also be useful. Be sure to rotate the birds eye view all four times, sometimes a view from one direction will be much high res, or newer, than the view from another. And sometimes, you just get a processed view of the same overhead shot. In any case, try them all. Generic1139 (talk) 21:46, 12 November 2015 (UTC)

Thanks, I didn't know about the bird's eye view - seeing it from an angle can really help. But I was trying to get the location of Taylor Hall (Hawkinsville, Georgia) and that was not sharp enough to tell. However, the western 3/4 of that block goes with the house, and I only see one thing that could be the house, so I think that must be it. (I did not get to see it when I was there - the address and the lat/long together were not enough - but it is way back with no good view from public property.) Bubba73 ^{You talkin' to me?} 00:38, 13 November 2015 (UTC)

Princeton (NJT station)

Help is needed for Princeton (NJT station). Details are at Talk:Princeton (NJT station). The station is mentioned in state documents are part of an HD, but the HD boundaries don't include it. Because it was marked as part of an HD, it may not have been acted on my the NPS in a thematic nomination. The article keeps bouncing back and forth between on the NRHP but with no refnum, or contributing property. Comments are welcome. Generic1139 (talk) 22:09, 16 November 2015 (UTC)

Refnum in infoboxes

Due to a prompt on Template talk:Infobox NRHP about an issue with the recent addition of links to the new Focus, I've now added code to the infobox that automatically places an article in Category:NRHP infobox needing cleanup if the refnum parameter does not start with a number. The infobox previously only added this category if the refnum parameter was completely blank. I thought this would only affect the ~50 articles mentioned on the talk page that came up on a search, so it would be easy to clean up, but as the job queue catches up, the category is becoming more and more populated. From a first check it seems many sites in Texas were bypassing the refnum check by just putting a generic NRISref in the refnum parameter, and there are several other random misuses of the parameter. Would anyone like to help cleaning up these articles? Currently there are 67 pages in the category. Thanks!--Dudemanfellabra (talk) 08:10, 17 November 2015 (UTC)

I suspect that Elkman's infobox generator may have at some time produced infoboxes without refnums (but with {{NRISref}} citation in the refnum field). Magic♪piano 12:25, 17 November 2015 (UTC)

I've been going through the original set with errors, and the cleanup category. It would be less tedious if NRISref had a way to handle dates past the old cutoff of July 9, 2010, otherwise, it is painful to come up with a weekly list reference. Now that focus returns meta data for entries will into 2014, NRISref or something similar should be able to help out by giving a reference to the NPS focus database. Generic1139 (talk) 13:55, 17 November 2015 (UTC)

Ouch. To do my bit for the cause, I started on the bottom of the alphabetical list, with the William J. Bryce House in Ft. Worth, Texas. This was created in April 2012 as an NRIS-only substub; 2 1/2 years later, an IP editor pasted the text of a historical marker verbatim into the middle of it. Between fixing the copyvio and trying to turn it into an article that a reader might actually find useful, I wound up spending about an hour on it. Be forewarned that if my limited sample is any guide, these articles will have lots more problems than just a missing refnum. — Ammodramus (talk) 14:03, 17 November 2015 (UTC)

Agreed. This is how I stumbled into the Princeton (NJT station) can of worms. Generic1139 (talk) 15:08, 17 November 2015 (UTC)

How to handle Dinosaur National Monument? Change NRHP infobox to protected area (similar to Buck Island Reef National Monument ? Or add the MRA reference number 64000073 as the NRHP refnum? Generic1139 (talk) 15:46, 17 November 2015 (UTC)

Dinosaur NM doesn't appear to actually be listed as a separate entity. It's not in Focus as a separate listing or in the old NRIS database. I'd change the infobox to protected area. Magic♪piano 16:38, 17 November 2015 (UTC)

Here's another weird one: what should we do about Swedish Farmsteads of Porter County, Indiana? The main article has an infobox for the "Swedish Farmsteads Historic District (Pending)" and it links to several other articles that have either no refnum or a very unusual-looking one, none of which appear to be actually listed on the Register. Charles Johnson Farm is supposedly a contributing property (it doesn't say to what historic district) with no refnum, and Anders Kjellberg Farm, Gust Lindstrom Farm, and Pete Larsen Farm have refnums that are longer than 8 digits (and in one case is hyphenated). They all cite some sort of National Register nomination without linking to it; the nominations are usually dated 1994 or 2008, but there's no evidence that any of these places were ever listed. (The properties themselves seem to be real, just not any of the NRHP info about them.) Aside from probably gutting any NRHP-related info from these articles, I'm at a loss as to what we should do with them. TheCatalyst31 ^{Reaction•Creation} 04:29, 18 November 2015 (UTC)

It gets worse. I tried Googling for one of the distinctive names in the purported Swedish Farmsteads HD, "Anders Kjellberg", and discovered that we've got two articles on the property: Anders Kjellberg Farm, which makes an apparently spurious NRHP claim, and Chellberg Farm, which doesn't, though it cites a nom form for a "Swedish Farmsteads" MPS (with no link in the citation). Worse still, both articles were created by the same editor, "Chellberg" in 2008 and "Kjellberg" in 2015. (This editor, Chris Light, also created the other articles in TheCatalyst's comment.)

Digging deeper, I find an online copy of the NPS's 2000 Cultural Landscape Report: Chellberg Farm. On p. 3 (which is p. 10 of the PDF), it states that the farm "is included in a Swedish Farming District that is potentially eligible for listing in the National Register of Historic Places". A footnote on p. 6 (p. 13 of the PDF) states that "[a] multiple property nomination is being prepared for the district by the Lakeshore historians".

Like TheCatalyst, I'm at something of a loss. My inclination is to leave the creator of the articles a note, asking that the Kjellberg/Chellberg articles be merged, that any claims of NRHP listing be sourced or deleted, and that a serious effort be made to find online sources so that other editors can check any dubious assertions. Thoughts? — Ammodramus (talk) 14:13, 18 November 2015 (UTC)

The note you left on the authors page sums it up nicely. Certainly, removing the claims that the properties are listed and removing the infobox seems non controversial. Claims of pending need to be backed up, but even pending doesn't mean it's on the list as had been discussed here many times. Generic1139 (talk) 18:07, 18 November 2015 (UTC)

Thanks to everyone for helping out. Sorry I haven't been able to participate as much as I usually do lately. Now that it appears we have taken care of all of the infoboxes that have no digits whatsoever, I have edited the infobox to include pages that have a refnum that has fewer than 8 digits, another error. I expect the cleanup category to populate shortly. If anyone would like to work on these, feel free to do so! Thanks again!--Dudemanfellabra (talk) 04:46, 20 November 2015 (UTC)

National Register of Historic Places in Bergen County, New Jersey

Part 1

Does anybody know the guy described in this article? There's a pretty good chance that he's done something here as well, and I think he could help avoid all the mess that follows below. I've got an (old?) address from the white pages but an e-mail or personal contact would help. Smallbones_(smalltalk) 18:37, 18 November 2015 (UTC)

Part 2

Bergen County, NJ has the strangest set of NRHP sites of anywhere I've seen. Consider that one Multiple listing nomination covers 190+ sites. That's more than Minneapolis and all of Hennepin County, MN. There about 280 sites total in a pretty rural area just NW of New York City. It gets stranger. There are at least 18 sites named the Ackerman-Something House, 2 of which are just "Ackerman House." There are about a dozen other family names with multiple houses named just like this. There are also different streets with the same names (and same family names) spread around the county, not to mention streets that change their names in the middle. And many of these colonial Dutch stone houses seem to be partially buried beneath modern additions. And we have confused some info in our articles and tables. Without criticizing anybody, that confusion seems inevitable. Did I mention winding hilly roads with no parking and lots of no trespassing signs?

Roy at Commons is plowing away at this, doing his usual very thorough job, but says he could use some help. Is there anybody who has experience with something like this? anybody know the area? @Jim.henderson, Daniel Case, and Nyttend: Smallbones_(smalltalk) 18:37, 18 November 2015 (UTC)

"a pretty rural area just NW of New York City." Really? Have you actually been there? Speaking as someone who was a) born there and b) lives pretty close to it even today, I consider it pretty suburban (only in the very northwestern part does it remotely get rural). It was well-settled at the time of the Revolution ... doesn't surprise me there's so many listed properties. Daniel Case (talk) 07:10, 19 November 2015 (UTC)

Hey! I found a nearby expert! Any chance you could wade thru the nom and help Roy a bit. BTW, I should have written "a fairly non-urban area just NW of NYC" instead of "a pretty rural area." These things are all relative of course. Smallbones_(smalltalk) 15:16, 19 November 2015 (UTC)

Might be a good thing to work on over Thanksgiving ... what specifically does he need help with? I have photographed a few of these Bergen County old Dutch stone houses in the past (Bergen County was particularly hard hit by New Jersey's early 20th century "boroughitis", so it has lots of fairly small communities (small even by the standards of elsewhere in northeastern NJ's Southern California-rivaling suburban sprawl, and each one of these seems to have ended up with at least one of those old houses) but nowhere near even a good chunk. Daniel Case (talk) 18:17, 20 November 2015 (UTC)

Roy very carefully researches most of the sites he photographs, e.g. using our articles+, and checking cords. Some of our articles are corrupt, likely due to "corrections" so they are throwing him off. No wonder since there are many names like Haring-Blauvelt-Demarest House, Haring-Blauvelt House, Blauvelt House 1, Blauvelt House 2, Blauvelt House 3, Blauvelt-Demarest House, and Cairns-Whitten-Blauvelt House (7 total), or see the 1st eight entries at National Register of Historic Places listings in Saddle River, New Jersey for just some of the Ackerman houses. So checking our articles against local knowledge would make a lot of sense. Maybe connecting him to somebody else, e.g. a historical society, would help as well.

As far as actual photography or traveling - going through some of the neighborhoods and seeing "do our photographs actually make sense?" might help and taking some pix yourself - it might help to wait until all the leaves fall. Contacting the guy in Part 1 might help. I know I can do it, but he might believe it more if the request comes from a local. I'll send you an e-mail. Smallbones_(smalltalk) 18:57, 20 November 2015 (UTC)

Part 3

We *had* very many photos of these sites but they were deleted. They are from the guy in part 1, on the "Historical Marker Database" www.HMdb.org (just search for Bergen County there). HMdb has freely licensed photos, except there's an additional requirement for commercial use that you can't download more than 25% of the photos in 1 area (e.g. in any municipality or county). That exception might seem impossible for us to deal with, except that, for all the sites I've seen in Bergen County, the guy in Part 1, has posted more than 4 photos. So if we limited ourselves to 1 photo from each site, mathematically we'd have to be below 25% in any area. Yes, there would be some additional issues to consider, like how do we make sure that only 1 photo per site isn't uploaded by multiple editors. The photos were deleted at Wikipedia:Possibly_unfree_files/2014_October_17 by folks who didn't seem to understand the issues or consider the math (on both sides).

I'm throwing up my hands here and just asking "what can we do?" Any help appreciated. Smallbones_(smalltalk) 18:37, 18 November 2015 (UTC)

Multiple Property forms

I believe the following statements are true about TR, MPS, MRA, and the like.

• Simply appearing in one of the multiple property forms does not confer "listed" status on a property. It must be in an HD as a contributing property, or be listed individually. An "Entered in the National Register" notation means the property will have its own reference number (sometimes given in the MPS) and its own nomination (sometimes a simple list with a copy of the MPS), but it will appear in the focus database.

• Articles about such lists do not get an NRHP infobox. They might get a project banner on the talk page, with a related importance, or if it contains a list of properties, some of which are otherwise listed, a list class.

• Articles about a property were the only NRHP justification is a mention in an MPS, MRA, TR, etc., and where the property does not have a Entered in the National Register stamp should be corrected. If the property is otherwise listed on the NRHP, provide the correct reference. If the property is not lsited, correct the article and remove the NRHP decoration.

• A multiple property form does not create a historic district.

• A pending property never gets an infobox.

I plan to correct the obvious cases in Category:NRHP infobox needing cleanup based on the above. Generic1139 (talk) 14:28, 19 November 2015 (UTC)

I think most of the above is correct, although I would disagree with your second point. I'm not sure why articles about listed NRHP properties would not get a NRHP infobox. See, for example Iron County MRA, where I believe most of the properties included are not notable enough to deserve their own article (individual NRHP listing notwithstanding). The infobox seems to me to add some context. Andrew Jameson (talk) 16:35, 19 November 2015 (UTC)

Some context, but not much - the reference numbers aren't complete and therefore not machine readable, the dates aren't specific as to what which listing they apply to. There are other ways to host the map and collage. If a reader wanted to click to get the NRIS data, they would need to go to the county list to find the reference number for a property. As a separate issue, it seems that almost all of the references are now dead links. Generic1139 (talk) 18:19, 19 November 2015 (UTC)

I've fixed all but Charles Johnson Farm and Swedish Farmsteads of Porter County, Indiana to give Chris Light a chance to respond to Ammodramus. I also plan to remove the infobox from Iron County MRA unless there is a consensus that it should be kept. Generic1139 (talk) 22:01, 19 November 2015 (UTC)

I agree with the other points, but I also disagree with the second point. I don't see a good reason why articles about multiple property submissions shouldn't have their own infoboxes, since an infobox can be a useful visual aid and a quick reference for information on a MPS. I'm especially concerned that we're doing this so soon after the infobox code change that put a bunch of MPS articles in Category:NRHP infobox needing cleanup; it seems like we're changing our article content to fit our new infobox code rather than the other way around. We should either change the code to find some way to work around MPS infoboxes or agree to some way to add reference numbers to MPS infoboxes (either in the style of the Iron County article, or by using the MPS reference numbers). TheCatalyst31 ^{Reaction•Creation} 04:48, 20 November 2015 (UTC)

I don't see the use for a set of partial reference numbers as in the Iron County article. They aren't correlated to a property, you can't search for them, what's the point? Also, this isn't an issue with the new infobox code - infoboxes without reference numbers have shown up in the cleanup category for quite a while. The new code simply catches those that were missed before because they had a reference in the otherwise empty refnum parameter. I'm fine with having an multiple listing flag or special nrhp_type, and using the MPS number (which always starts with 64, not the year), as long as we don't try to cram 70 reference numbers into it. If we list properties in an MPS article, I'd also argue that they should use the NRHP Row format as well - easier to maintain, and would look more like other list articles - just with a longer intro and the descriptions in the table with the image, location, and usable refnum. Generic1139 (talk) 06:45, 20 November 2015 (UTC)

I have no particular objection to changing or reformatting the refnums, if that's what's causing heartburn. However, I think that a table format wouldn't work well for this particular article (although it probably would work well for most MPS articles). Most of the properties which are part of the Iron County MRA are likely not notable enough to deserve their own article, so the group article is the de facto repository of information for those particular properties. That makes the descriptions longer than will easily fit in "description" block of the table. Andrew Jameson (talk) 12:05, 20 November 2015 (UTC)

As currently structured, the NRHP infobox is a container for a historic place, using one of the defined types.

Accepted values

"nhl" — National Historic Landmark
"hd" — Historic District (more information here)
"nhld" — National Historic Landmark District (NHL that is also a historic district)
"cp" — Contributing Property to a Historic District (more info here)
"nhldcp" — Contributing Property to a National Historic Landmark District
"indcp" — Contributing Property that is also individually listed (more info here)
"nmon" — National Monument
"nhs" — National Historic Site
"nmem" — National Memorial
"nhp" — National Historic Park
"nb" — National Battlefield
"nmp" — National Military Park
"nbp" — National Battlefield Park
"nhs" — National Historic Site
"nhr" — National Historic Reserve
"ihs" — International Historic Site

If a property has been delisted from one of these designations, append "former" to the beginning of the value (i.e. "formernhl", "formerhd", etc.) "formernrhp" is also available, though not required. For more information about delistings, click here.

We can:

• Make a new nrhp_type for MPS (optionally, each type of MPS) for the existing infobox. This one might be the best choice, but it would be the only type that isn't a specific listed property.

-or-

• Make a new info box altogether

We need a convention for listing multiple reference numbers for MPS. I suggest we avoid that problem and simply use the MPS reference number, and if the article has table of listed properties, or a series of summary paragraphs, then the table or paragraphs should give the refnum, and use something similar to [{{NRHP Focus|refnum|url=y}} refnum] to make them clickable.

We also need to agree that properties that are listed on the MPS, but were not added individually or are part of an HD, don't get an infobox with a CP and a partof_refnum = the MPS.

I suggest we keep (and continue to refine) the validity checks on refnum, as that process finds far more problems than it does false positives. Generic1139 (talk) 16:12, 20 November 2015 (UTC)

And, FWIW, with a quick look at the 68 articles in Category:National Register of Historic Places Multiple Property Submissions and its sub categories that are actually MPS articles, 43 use a single refnum (usually the MPS), 8 used multiple refnums, 7 have no infobox. I'm modifying my changes to use the MPS refnum for in the cases where the refnum field was empty or non-numeric (for now).

New listing today that may be is a first of its kind

In today's newly-announced listings there's one from Washington County, Arkansas, that I can't wait to see a picture of:

Prairie Grove Airlight Outdoor Telephone Booth, SW. corner of E. Douglas and Parker Sts., Prairie Grove 15000291

Would this be the first time a telephone booth has been listed on the NRHP? I know Britain lists just about every one of those old red phone booths they can find, but I don't think we've ever granted one this honor (Not that we shouldn't ... they sort of are disappearing, and should one day arouse the same curious interest that mounting blocks do today). Daniel Case (talk) 18:24, 20 November 2015 (UTC)

Addendum: Yes, apparently it is the first. OK, someone needs to get a picture so we can start an article about it and get it nominated to DYK. Daniel Case (talk) 18:31, 20 November 2015 (UTC)

Photograph here, for people who aren't old enough to remember them. Nomination here (with lots more pictures). This one should be fun to write up. Magic♪piano 18:35, 20 November 2015 (UTC)

I have pinged Valis55, who has done yeoman's work photographing many of northwestern Arkansas's listings. Definitely a DYK. Magic♪piano 18:39, 20 November 2015 (UTC)

I called the phone number given for it in the nom, sorry it was busy. Don't ask me why I called the phone number - it was there and why not? Which raises the question of whether we should put the number in the article. We might ask the folks at the motel for a quikpic. Smallbones_(smalltalk) 19:14, 20 November 2015 (UTC)

This also raises the question of whether there are other famous phone booths (not numbers or simple phones, but booths). Maybe in Metropolis, may something to do with the Telephone booth stuffing craze? Smallbones_(smalltalk) 20:20, 20 November 2015 (UTC)