Wikipedia talk:WikiProject National Register of Historic Places/Progress/Archive 1
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 |
Color scheme
What color scheme scale is being used for the maps? 25or6to4 (talk) 02:16, 2 April 2013 (UTC)
- There is an ongoing discussion at WT:NRHP about this. Please feel free to contribute. When the map is next updated, a key will be included.--Dudemanfellabra (talk) 02:29, 2 April 2013 (UTC)
This was probably discussed a long time ago, but to me red indicates where action needs to be taken - add photos, create articles, etc. But red is used in the opposite way. Bubba73 You talkin' to me? 15:21, 12 June 2015 (UTC)
Moving forward
User:Smallbones has brought up some good ideas at WT:NRHP, but since they are specific to this page, I'll move the discussion here. His list of ideas is copied below:
- We need a strategy/procedure for updating this - say every 2 months
- We should use this as a motivator, e.g. by placing the pix and a notice on the state projects pages saying something like "WP:NRHP needs your help. All those counties colored blue in (your state) need photos and articles for the listings on the NRHP located there." (needs work)
- We should figure out a way to help out editors who want to concentrate on the blue states - maybe a State of the Month to concentrate our article writing efforts in one state (say Kentucky for May), which will also help the photographers in that state find the sites more easily.
As to #1, I agree.. two months sounds like a good interval to update the image. This page itself can be updated whenever--as close to real time as possible--and the image script can be run every two months to generate SVG output. I envision people updating this page every time they update the /pictures and the /articles subpages. Maybe we should drop a note there to also update the statistics here. There is also User:AlexNewArtBot/NRHPSearchResult, which catches many new articles (but not pictures) that aren't added to the /articles subpage. Someone could check that every few days and update stats here. But that isn't to say we shouldn't still do what we just did and sweep through all the states at once. Doing it that way makes for fewer errors, like double counting a new article/image, which may be introduced easily if more than one person is keeping track of everything. And then of course there is the possibility of extending the scripts in place to become fully automated so that this page is constantly updated, and no actual editors have to worry about it. That's IMO the best option, but also the toughest to implement.
For #2, I like the idea of motivating state projects to get involved, although I think most of the editors that are members of those also edit here. A notice wouldn't hurt, though. We could even tie it together with #3 and only put a notice on a state when it's their turn to be state of the month. That way we wouldn't just spam all the states at once and turn people off.
Something Smallbones didn't mention but that I thought of after this drive was that we could split off each state into its own supbage, e.g. Wikipedia:WikiProject National Register of Historic Places/Progress/Alabama and allow editors in that state to discuss progress. The table itself could be transcluded back onto this page, but state-specific discussion should be on the subpage inside noinclude tags, kind of how the notes on the /articles and /pictures pages work. The section titles here could then be changed to link to the subpages instead of the NRHP list. I like the list link being there, though, so maybe we could also change the links in the table to point to the state list instead of the article about the state itself.
Does anyone else have any other ideas? I like where this is going!--Dudemanfellabra (talk) 19:22, 8 April 2013 (UTC)
- Do we have state maps by counties? They should be easy to produce since they use the same info, and the counties would be visible.--Ymblanter (talk) 19:41, 8 April 2013 (UTC)
- DMFB, if you're not surfeited with compliments, let me offer another. The main project page looks great with the two maps on it.
- Unfortunately, there's a rather dark cloud to go with this silver lining. The article map is going to create lots of positive reinforcement for the creators of robo-stubs, as they happily watch county after county turn red. Worse, the more infobox-and-two-sentences stubs are out there, the more likely it is that new editors will believe that that's all they need to do. I'm afraid that we'll wind up with lots of "stub deserts" on the map: big red patches where all the articles are "$SITE is a historic site located in $LOCATION. It was built in $YEAR1 and added to the NRHP in $YEAR2."
- Is there anything we can do about this? The only possible remedy I can think of is a rather extreme one: ask Elkman to take down the infobox generator, or at least to hide it at a URL that'd only be revealed to editors who've demonstrated that they're committed to creating articles of substance. I'd hate to deprive other editors of a useful tool, though, and I'm not even sure if it'd work: I don't know how difficult it'd be to write a new infobox generator, and once it's written the torrent of substubs can be set to flowing again. Any better ideas? Ammodramus (talk) 20:34, 8 April 2013 (UTC)
- One possibility would be WikiWork, or something similar, as shown in this map for the US roads Wikiproject. That way, we could maybe have two maps, one showing articles not written, and another for counties suffering from many stubs. Chris857 (talk) 20:47, 8 April 2013 (UTC)
- Elkman's infobox generator is not to blame for the substubs. He's quite clear about what you're supposed to do with it, and it's still really easy to write a substub without it. Maybe an enforceable rule is the solution. If we ever get around to those "focussed, structured discussions" that Arbcom recommended, we could set an acceptable minimum article standard. Ntsimp (talk) 22:12, 8 April 2013 (UTC)
- @Ymblanter: We don't have state maps, but that's something we could do. If we decide to go the subpage route, each state's subpage could include the state map.
- @Ammodramus: I agree that this has the potential to encourage sub-stubbers, but that's just kind of something we'll have to look out for. We have notes in various places saying not to use the Elkman generator to create sub-stubs, but it happens and will continue to happen. Even so, I think taking the generator down or limiting its use will have a net negative effect.
- Something I did think of long term that could help with that problem was making maps similar to this for average quality of articles (i.e. Stub, Start, GA, etc.) in each county. That would be a considerably harder list to compile, though, and it would definitely require a bot. I think just the existence of the articled map here is a good start toward that final goal. We also have on the main project page a collection of assessment statistics, but they aren't broken down by geographical area. There is a lot that can be done with statistics, but until we can get some kind of bot to do it for us, we're really limited. I mean even with the scripts for doing this, it still took a good two weeks to get everything done.. and that's just tallying up numbers.. no kind of detailed analysis like getting the class of articles would be. That would take way longer.--Dudemanfellabra (talk) 22:14, 8 April 2013 (UTC)
- The problem with using WikiWork or any other kind of assessment-based analysis for NRHP articles is that it doesn't distinguish between two-line substubs and the more detailed articles that editors like Pubdog write, which often get assessed as stubs. The reason the WikiWork system works for projects like USRD is that it has a very well-defined assessment system (which in the case of USRD is mainly possible because almost every road article has the same three sections). If we were to apply something like WikiWork to NRHP, we'd need a better set of assessment criteria; for instance, we might want to bump the longer stubs up to start-class and the longer start-class articles up to C-class. Unfortunately, reassessing every article is going to take much longer than anything that's been proposed, but otherwise an assessment-based analysis won't tell us anything useful. TheCatalyst31 Reaction•Creation 23:22, 8 April 2013 (UTC)
- [Written before Smallbones put in the new section] I agree with Ntsimp's comment — you can write most of your substub from nationalregisterofhistoricplaces.com, and basically everything that's missing, like coords, can be found at our county lists; what's more, most of the information is easily found by downloading the NRIS anyway. What's more, substubs are responsible for much of Ohio being really really sparse — I've done lots of writing in Ohio, but you wouldn't notice (on the map) if most if it hadn't happened, because I've put more work into expanding NrhpBot stubs in the Cincinnati area from this to this. This is a big reason I spend more time getting photos than writing articles — as long as your photo shows the spot accurately, it's sufficient, while you can spend tons of time writing and writing on the same article without there being a single way of saying "This is sufficient". On top of all of those, PD or freely licensed photos are much more helpful to reusers: anybody who's researching an NR-listed property will be able to get information from the nomination form, but they might not be able to get a useful photo from a specific time period. Nobody will be able to get a replacement version of File:The Chadwick, Indianapolis.jpg, but if they want to research it they'll have an easy time. Note that none of this is at all aimed at people who prefer writing articles; I'm purely talking about what I like to do. Nyttend (talk) 00:07, 9 April 2013 (UTC)
- TheCatalyst31 makes some good points. I, too, think it's horrible that a one-sentence stub would be grouped in with Pubdog's and others' "good" stubs. I agree that either we need to have a project-wide consensus on the minimal acceptable stub (which we've tried several times and never got going) or, alternatively, we just add a new assessment level "Sub-Stub". Yes, that's kind of derogatory, but it's true. If you're creating one sentence stubs, you deserve to be derogated.--Dudemanfellabra (talk) 17:37, 9 April 2013 (UTC)
- I like the "sub-stub" idea myself, since it allows us to differentiate the two kinds of stubs we have now without having to redefine all the higher assessment classes. I don't see it as much more derogatory than "stub", especially since articles used to be tagged as substubs; of course, if that name is just going to start another nasty and unproductive discussion, we can easily separate the really short articles in some other way. I don't entirely mind if people want to start these articles, but they should at least say what they are so other people know they need to be expanded. TheCatalyst31 Reaction•Creation 23:46, 9 April 2013 (UTC)
- TheCatalyst31 makes some good points. I, too, think it's horrible that a one-sentence stub would be grouped in with Pubdog's and others' "good" stubs. I agree that either we need to have a project-wide consensus on the minimal acceptable stub (which we've tried several times and never got going) or, alternatively, we just add a new assessment level "Sub-Stub". Yes, that's kind of derogatory, but it's true. If you're creating one sentence stubs, you deserve to be derogated.--Dudemanfellabra (talk) 17:37, 9 April 2013 (UTC)
- [Written before Smallbones put in the new section] I agree with Ntsimp's comment — you can write most of your substub from nationalregisterofhistoricplaces.com, and basically everything that's missing, like coords, can be found at our county lists; what's more, most of the information is easily found by downloading the NRIS anyway. What's more, substubs are responsible for much of Ohio being really really sparse — I've done lots of writing in Ohio, but you wouldn't notice (on the map) if most if it hadn't happened, because I've put more work into expanding NrhpBot stubs in the Cincinnati area from this to this. This is a big reason I spend more time getting photos than writing articles — as long as your photo shows the spot accurately, it's sufficient, while you can spend tons of time writing and writing on the same article without there being a single way of saying "This is sufficient". On top of all of those, PD or freely licensed photos are much more helpful to reusers: anybody who's researching an NR-listed property will be able to get information from the nomination form, but they might not be able to get a useful photo from a specific time period. Nobody will be able to get a replacement version of File:The Chadwick, Indianapolis.jpg, but if they want to research it they'll have an easy time. Note that none of this is at all aimed at people who prefer writing articles; I'm purely talking about what I like to do. Nyttend (talk) 00:07, 9 April 2013 (UTC)
- The problem with using WikiWork or any other kind of assessment-based analysis for NRHP articles is that it doesn't distinguish between two-line substubs and the more detailed articles that editors like Pubdog write, which often get assessed as stubs. The reason the WikiWork system works for projects like USRD is that it has a very well-defined assessment system (which in the case of USRD is mainly possible because almost every road article has the same three sections). If we were to apply something like WikiWork to NRHP, we'd need a better set of assessment criteria; for instance, we might want to bump the longer stubs up to start-class and the longer start-class articles up to C-class. Unfortunately, reassessing every article is going to take much longer than anything that's been proposed, but otherwise an assessment-based analysis won't tell us anything useful. TheCatalyst31 Reaction•Creation 23:22, 8 April 2013 (UTC)
As TheCatalyst31 predicted, there has been some talk of creating stubs (although I would put them closer to Pubdog's quality stubs than sub-stubs) just to turn counties on these maps dark red. Sigh. (For reference, here is the link he was trying to type.)--Dudemanfellabra (talk) 23:21, 15 April 2013 (UTC)
- Conversely, writing articles for upcoming NRHPs might be considered. There's a gap of one to two months between a site showing up as pending and it being approved. And before they go national, they have to go through a state approval process. At least they do in Florida and Georgia. Some sites are notable without the NRHP status, and I would think those articles could be written straightaway. With appropriate sources, of course. Otherwise, they could be pre-written and then the articles created once approved. Maybe a workspace could be created for making pending NRHPs? Especially now that the nomination info seems to be being uploaded for new listings concurrent with their approval. Hey, anything to get ahead of the curve. :) --‖ Ebyabe talk - Border Town ‖ 18:09, 18 April 2013 (UTC)
Feedback, standards, projects, etc.
Ammo is right that this can give a lot of positive feedback for folks who just want to write super-stub articles. I suppose the same might be said (and aimed at folks like me!) about the illustrated maps "This will give a lot of positive feedback for folks who want to just take snapshots instead of quality photos." This is probably less important than the article map, and such a criticism wouldn't affect me at all - I enjoy taking my snapshots and if somebody else wants to take better photos - more power to them. I suppose the best thing to do is to emphasize that we would appreciate additions of stubs of a certain quality (we better define this) but that a 2 line stub will just make the map a less powerful tool.
I took the state table we have now and found the "top ten" for articles needed (number, not percentage) and photos needed (also raw number). The states that are on both lists are Kentucky (top of both lists - the rest are in no special order), Texas, Arkansas, Virginia, Missouri, Georgia. I strongly suggest that we get something going for Kentucky. Of course, there's not much most people can do in taking photos, unless they live in or near Kentucky, so most of our effort would be in writing articles. Could we make a worthwhile difference in a month? Well if we got 10 WP:NRHP editors to write 20 articles each, we'd probably get just enough articles written to put KY into 2nd place behind Texas. The real difference would be if we got editors in KY involved, and that would pay dividends every month.
I don't think that the "articles needed, photos needed" method would always be the best selection method to use for a target state. We clearly need something to get people started in the great plains/rocky mountain states. Ammo has shown that just 1 editor in a state can do wonders. I wonder if he has any suggestions on how to find the 1 or 2 editors we need in each state in the area (well, Texas may need 5 or 10 just because of the distances involved).
All the best,
Smallbones(smalltalk) 23:51, 8 April 2013 (UTC)
- I've done my best with Kentucky, having photographed about 1/8 of its sites so far :-) You picked a good time to make this statement; I'm planning on taking Thursday through Saturday this week for a trip getting photos in Little Egypt, the Jackson Purchase, the Bootheel, western Tennessee, and the northeastern corner of Arkansas. I don't know about articles, but a lot of our Kentucky photography has been a now-inactive editor in the Cincinnati area, Greg5030. If only we could convince Bedford to come back; he's in metro Louisville and is well situated to get tons of sites in the Bluegrass region, where most of the state's sites are found. Kentucky is slightly hard, however — they have a ridiculous number of hard-to-get sites because they've done tons of work with rural MPS projects, and of course lots of farmsteads are either gone or located at the ends of ½-mile-long lanes that you can't photograph. Take the Shelby County map, for example — to get all of those sites, you're going to be driving virtually every road in the county, while here in Indiana I've had an easier time (and even easier in far eastern Illinois) because there are far fewer sites in general and a much higher percentage of sites are in a few towns. More encouragement, I guess, for us to cover Kentucky by writing articles :-) Nyttend (talk) 00:18, 9 April 2013 (UTC)
- Kentucky is quite nice in the spring I've heard. dm (talk) 01:26, 9 April 2013 (UTC)
- I prefer taking photos myself and I certainly don't mean any disrespect to anybody who has taken photos there. It just seems that there must be something we can do to help out in KY and some other states if we don't live close enough to take pix ourselves. Perhaps we can search through HABS or Carol Highsmith's collection, or ... Writing articles (better than substubs) can't hurt. Enjoy the trip! Smallbones(smalltalk) 03:47, 9 April 2013 (UTC)
- I'm sorry I came off that way — I saw your statement as saying (among other things) "We've gotten as many photos as possible, but most of us can't help at all that way". I never imagined that you were complaining about what I've done or not done! I don't at all object when people write articles, but as I said above, I personally prefer doing the photos, if for no other reason than that photos are much more time-dependent; photos like File:Drexel School 2.JPG are critical. Unfortunately, there's not a ton of HABS that can be done for Kentucky; some years ago I uploaded lots of HABS images for the state, and I believe I uploaded everything that seemed relevant. Surely there's more that can be done, whether uploading things that I overlooked or uploading images that weren't online a few years ago; and I never checked Highsmith. But yeah, a lot of Kentucky is like Chester County PA, with the exception that much of it isn't as old or as interesting; a lot is dumpy places with poor maintenance, although you do find some interesting stuff. Nyttend (talk) 04:01, 9 April 2013 (UTC)
- For me it was a "Rough day at the office." No apology should have been needed. But thanks anyway. Smallbones(smalltalk) 18:27, 9 April 2013 (UTC)
- Wacky suggestion, perhaps, but consider contacting special interest groups that travel? Like if you know folks that go to science fiction conventions, or are medieval or civil war reenactors. I've been thinking about seeing if any of my old SCA friends might be interested. Wouldn't work for article-writing, but maybe we could get some extra roving photographers if we're lucky. --‖ Ebyabe talk - State of the Union ‖ 18:19, 18 April 2013 (UTC)
- For me it was a "Rough day at the office." No apology should have been needed. But thanks anyway. Smallbones(smalltalk) 18:27, 9 April 2013 (UTC)
- I'm sorry I came off that way — I saw your statement as saying (among other things) "We've gotten as many photos as possible, but most of us can't help at all that way". I never imagined that you were complaining about what I've done or not done! I don't at all object when people write articles, but as I said above, I personally prefer doing the photos, if for no other reason than that photos are much more time-dependent; photos like File:Drexel School 2.JPG are critical. Unfortunately, there's not a ton of HABS that can be done for Kentucky; some years ago I uploaded lots of HABS images for the state, and I believe I uploaded everything that seemed relevant. Surely there's more that can be done, whether uploading things that I overlooked or uploading images that weren't online a few years ago; and I never checked Highsmith. But yeah, a lot of Kentucky is like Chester County PA, with the exception that much of it isn't as old or as interesting; a lot is dumpy places with poor maintenance, although you do find some interesting stuff. Nyttend (talk) 04:01, 9 April 2013 (UTC)
- I prefer taking photos myself and I certainly don't mean any disrespect to anybody who has taken photos there. It just seems that there must be something we can do to help out in KY and some other states if we don't live close enough to take pix ourselves. Perhaps we can search through HABS or Carol Highsmith's collection, or ... Writing articles (better than substubs) can't hurt. Enjoy the trip! Smallbones(smalltalk) 03:47, 9 April 2013 (UTC)
- Kentucky is quite nice in the spring I've heard. dm (talk) 01:26, 9 April 2013 (UTC)
- On the article-writing side of things, until recently Kentucky had nearly nothing online in terms of NRHP resources, so if anyone who wasn't from the area wanted to write more than a two-line stub about some of these places they would have to do significant research. This has finally changed for Kentucky since the NRIS posted its nomination forms, but it partially explains why it lagged so far behind for so long. If we could get some of the national editors to work on KY sites now it could probably do some good. On a related note, availability of resources is probably something we should consider if we decide to do a State of the Month type thing; if we choose a state like Georgia that doesn't have NRHP forms online or a decent state-maintained database, it's going to be a lot harder to start substantial articles. TheCatalyst31 Reaction•Creation 04:33, 9 April 2013 (UTC)
- Actually, they've had lots online for quite a while; during the summer of 2010, when I got a lot of the photos at the Jessamine and Mercer County lists, I wasn't able to find anything (other than address-restricted and new listings) that wasn't online. Basically, as far as I can tell, the big lag by Focus was updating their page listing the states that were already online. Nyttend (talk) 05:00, 9 April 2013 (UTC)
State of the month projects
I'd like to take advantage of the new tables/graphs while we're still thinking about it, to get something "more important" started. The tables/graphs are wonderful tools that can be used for many things, but ultimately the "more important" thing referred to above, is how we use the tools. Recruiting new editors and photographers to fill in the blue spots is the first use that I thought about but I'm sure we can list a few more. "Recruiting" might even be too strong a word - just letting the appropriately-located people know that we could use their help might be a better way of looking at it.
For Kentucky - if everybody agrees this is a good place to start - we could almost start now and go to June 1. I think User:acdixon has a lot of interest in KY history and WP:KY, so I could ask him if he knows of anybody interested and if it would make sense to post to WP:KY. We definitely shouldn't force anything down anybody's throat. The main goal should be to find interested photographers in the area to take pix and other folks who are interested in the local history to write articles. WP:NRHP's role would be to help the new folks to work well on NRHP pix and articles, and to give the state project a kick start.
We'd then need to find a state, ideally each month, that editors and local photogs would be interested in. Spamming state projects would likely be counterproductive, but I think a notice at WP:USA with the graphs would likely attract a few interested folks, and personally contacting people you'd think would be interested would be the best way. All we'd need is a brief statement of goals - probably on a subpage of WP:NRHP or a subpage here, and a list of the next 3 or 4 states to be featured. If there are no other volunteers for June, I'll suggest New Jersey - not because it's a blue state, but because its got the highest number of 3 line NRHP stubs per square mile of any place in the country (educated guess- Focus just became the 1st online source of nominations).
Smallbones(smalltalk) 19:08, 9 April 2013 (UTC)
- Does it have more than Rhode Island or Massachusetts? Virtually everything in those states has an article, and most of the articles are "SITENAME is a historic building in PLACENAME, Massachusetts. On DATE, it was added to the National Register of Historic Places." Nyttend (talk) 22:00, 9 April 2013 (UTC)
- Yes, New Jersey is a good candidate for an article improvement campaign, because a) it's many stubs are just like what Nyttend asks, and b) the better NRHP nomination documents are now available. But, what do you get for making any improvement, can there be a tally and corresponding map for "improved articles"? Maybe we should define "Start" quality within this wikiproject to be a pretty good stub with NRHP nomination document included and with some explanation of why the site is deemed historically significant and a certain article length achieved. I.e. define Start to be very measurable, very objective. Then improvement campaigns would be intended to bring articles up to that level, and there could be a tally and graphic of these. Currently, MA and NJ and RI would show low percentages as those are mostly NRIS-only stubs; ND and UT and CT would be higher already or more easily improved as those mostly have NRHP documents; PA and NY and MD would be higher percentages already. It seems highly unproductive/unpleasant/uncollegial/etc to fight against anyone adding pics or starting articles at a low level, which is allowed by policy and guidelines and practice, and which supports higher level development (e.g. it solidifies the list-system, it protects supporting disambiguation, it allows display of larger-than-thumbnail pics, it appears to assist new editors who will contribute but are unwilling/fearful of the step of article creation, etc., etc.). Not everyone appreciates those purposes. But we all do share common ground of valuing higher level works. We should all agree that achieving a high percentage of "Start"-quality articles in a county is good. Define an NRHP-Start level to be achievable, objective, not too hard, and set up another tally column and another map graphic? --doncram 22:23, 9 April 2013 (UTC)
- To be honest, that sounds like a really good definition of a stub. Anything that doesn't include some explanation of why the site is historically significant shouldn't be created in my opinion. If you're going to take the time to create an article, at least put the five minutes of effort necessary to find out why that site is important. Anything that doesn't explain that is not even a stub.. it's a sub-stub. Instead of lowering the standards for start, why not just make "Sub-Stub" an assessment level, like I suggested above?--Dudemanfellabra (talk) 23:14, 9 April 2013 (UTC)
- Yes, New Jersey is a good candidate for an article improvement campaign, because a) it's many stubs are just like what Nyttend asks, and b) the better NRHP nomination documents are now available. But, what do you get for making any improvement, can there be a tally and corresponding map for "improved articles"? Maybe we should define "Start" quality within this wikiproject to be a pretty good stub with NRHP nomination document included and with some explanation of why the site is deemed historically significant and a certain article length achieved. I.e. define Start to be very measurable, very objective. Then improvement campaigns would be intended to bring articles up to that level, and there could be a tally and graphic of these. Currently, MA and NJ and RI would show low percentages as those are mostly NRIS-only stubs; ND and UT and CT would be higher already or more easily improved as those mostly have NRHP documents; PA and NY and MD would be higher percentages already. It seems highly unproductive/unpleasant/uncollegial/etc to fight against anyone adding pics or starting articles at a low level, which is allowed by policy and guidelines and practice, and which supports higher level development (e.g. it solidifies the list-system, it protects supporting disambiguation, it allows display of larger-than-thumbnail pics, it appears to assist new editors who will contribute but are unwilling/fearful of the step of article creation, etc., etc.). Not everyone appreciates those purposes. But we all do share common ground of valuing higher level works. We should all agree that achieving a high percentage of "Start"-quality articles in a county is good. Define an NRHP-Start level to be achievable, objective, not too hard, and set up another tally column and another map graphic? --doncram 22:23, 9 April 2013 (UTC)
- Per Ntsimp, above: "If we ever get around to those 'focussed, structured discussions' that Arbcom recommended, we could set an acceptable minimum article standard."
- How can we initiate this? Can we as a WikiProject set an enforceable standard for a minimum NRHP stub, one under which two-sentence NRIS-based robo-stubs are no longer "allowed by policy and guidelines and practice"?
- I'd take Doncram's "Start" definition as a minimal stub, as Dudemanfellabra suggests. However, I'd modify it by removing the requirement that it include a citation to the nom form. There are still nom forms that aren't available online (for instance, the Nebraska State Historical Society doesn't appear to be putting nom forms for AR sites on its website, and it looks like Focus might be doing likewise), but a more than adequate article might be achievable using other sources.
- I'd also go into a little more detail about "some explanation of why the site is historically significant", to make it clear that something like "The site is notable for its $STYLE architecture" or "...for its connection to politics/government" is insufficient.
- I assume that this discussion will have to be moved to the main project page. I'm afraid I know too little about the making of Wikipolicy to say much else: will it require, for instance, an RfC, or is there some even more formal procedure that has to be followed? Whatever it takes, I'd sugggest that we undertake it, before the proliferation of substubs renders DMFB's percent-articled map meaningless. Ammodramus (talk) 00:18, 10 April 2013 (UTC)
- Count articles meeting DYK-length How about define a "Start-type" standard for Wikiproject NRHP to be simply DYK-length (which is 1500 characters outside of infoboxes and tables)? The DYK-check tool makes this easy to measure for any one of us editing, and it is apparently programmable so a bot could probably identify and count all DYK-length articles. Getting 1500 characters is unlikely without having obtained an NRHP nom doc, or other serious sources, but whether or not the NRHP doc has been obtained doesn't have to be measured by this standard, so it is easy and objective to count articles meeting it. There is no way the NRHP wikiproject can define "Stub" differently than everywhere else; any low level article existing is a stub, it needs to be given a stub template, etc. Defining a "sub-stub" and going around labelling articles would be awfully negative. For wide usage in counting, the standard would need to be very quickly, easily measurable. So, how about setting a high-enough-to-satisfy-everyone standard for a higher, Start-type rating, i.e. just DYK-length, and then promoting that by adding a column, setting up a corresponding map reporting it by county, etc.
- A bot could identify whether there is a pic or not, whether there is an article or not, and probably also whether the article is a DYK-length or not, for automated counting by county every month, say. So graph "illustrated", "articled", and "DYK-length". While the first two are at about 50% on average with variation above and below, my guess is that "DYK-length" will be under 20% in every one of the U.S. states, to start.
- We could redefine "Start" for this wikiproject to mean exactly that, DYK-length, or not; or a DYK-length rating could be added to the project's assessment system as a another quality measure like "class=related" was done, before, i guess; the DYK-length quality could simply be measured by a bot and then indicate "DYK-length=Yes" or "class=DYK-length" on the Talk page of every article (for traditional project-wide tally-table reporting). --doncram 17:31, 11 April 2013 (UTC)
- That doesn't address the problem of distinguishing two-line articles from higher-quality stubs, though. For instance, take Leggett House (Merced, California), which I expanded yesterday. The original version is a two-line article that doesn't explain the house's significance beyond the architectural style, and the current revision is a "quality stub" (if I do say so myself) that describes the house's prominent architectural features and association with historical figures. However, both versions are under 1,500 characters, so under the system you suggest they would be assessed the same way, which is precisely the problem I wanted to avoid in the first place. TheCatalyst31 Reaction•Creation 22:28, 11 April 2013 (UTC)
- Yep. I happen to agree that there should be recognition of some "pretty good stub" or other name for articles that serve their purpose well. I happen to dislike many of the longer, padded-out DYK-length articles even. A few sentences with a link to an NRHP nom doc and links to other sources if available, is the proper encyclopedic treatment for many NRHP-listed places, in my opinion. However I think a "sub-stub" definition is a non-starter, is derogatory, there's a reason sub-stubs are no longer labelled. So, then, we have to use a positive definition "stub-plus" or "pretty good stub" or "good stub" or define this to be what we mean by "Start"-level. It would be easiest, administration-wise, to use this as "Start" and to begin to use "C-class" for what was previously Start, etc., as you suggested further up this page. But if there is too much negative energy here, too much fear of encouraging short articles, then maybe only defining a longer DYK thing that everyone agrees is better than a 2-line stub is the only way to advance for now. --doncram 01:06, 16 April 2013 (UTC)
- That doesn't address the problem of distinguishing two-line articles from higher-quality stubs, though. For instance, take Leggett House (Merced, California), which I expanded yesterday. The original version is a two-line article that doesn't explain the house's significance beyond the architectural style, and the current revision is a "quality stub" (if I do say so myself) that describes the house's prominent architectural features and association with historical figures. However, both versions are under 1,500 characters, so under the system you suggest they would be assessed the same way, which is precisely the problem I wanted to avoid in the first place. TheCatalyst31 Reaction•Creation 22:28, 11 April 2013 (UTC)
Maybe defining 750 characters as the level for "stub-plus" or "Start" would work? That is half the 1500 DYK size requirement, and is a pretty good length for good, short NRHP articles, like the improved Leggett House (Merced, California) article (at 895 characters and 2428 bytes now, vs. 235 characters and 1390 bytes before), seen easily by use of wp:DYKCHECK. Big question: would such a size criterion be programmable into the automatic updating of tables here? --doncram 12:42, 16 May 2013 (UTC)
- Legget House looks like a respectable, average, "this is how all stubs should be" stub to me. There isn't anything "plus" about it. In fact, maybe it would be good to set the line for an acceptable stub to be at 750 characters. Anything below that should be developed in userspace or some other non-mainspace location.
- As far as the size criterion, I'm not sure exactly what you mean. The script just checks if an article exists or not... nothing to do with assessment.--Dudemanfellabra (talk) 17:15, 16 May 2013 (UTC)
New validation code
After working on this basically all day yesterday, I have come up with some code that will validate a list based on the individual county lists. The script adds a button atop the page that reads "Validate county data." When the button is clicked, the script will go down the table(s), extract the links to the county lists, load each one in the background, pull out the stats (total listings, number illustrated, number articled, and percentages), and compare those to the ones on the list. If they don't match up, it colors the differing cell red and gives the script output as a tooltip so you can easily correct it. It can handle not only counties which have their own list articles but also counties which are sections on a different page. It does not (yet?) edit the page to update the information, partly because there are still some bugs to be worked out (or more likely worked around):
- If a county has its own page but the link to that list is a redirect here, the script automatically assumes the list it's looking for is a sublist (i.e. not on its own page), and when it doesn't find one, it returns "error" for all the quantities in that county (but keeps validating the other counties). To fix this, we need to change the links on this page to point directly to the correct article rather than going through a redirect. There are several links in Alaska that need to be changed to have an em-dash instead of a hyphen, and several city lists need to be updated to bypass redirects. What is displayed on the link can be changed, but the actual target needs to be a direct link.
- If the link in the table points to a section on a state or other page, the displayed text of the link needs to match exactly the section title on the state list. The script returns errors for that county's stats if the two don't match up. For example, some counties have spaces here ("La Salle") but don't have spaces on the state lists ("LaSalle"). These links need to be updated to display the correct county names.
- If a county has no listings in it, the script can only detect this if there is a section on the state page saying there are no listings in the county. If there is no such section, the script returns "error" for that county's stats instead of zeros. One way to work around this would be to gather a list of all counties which have no listings and no sections on their respective state lists. These counties can then be hard-programmed into the script, avoiding the error.
- Incomplete list below:
- National Register of Historic Places listings in Wade Hampton Census Area, Alaska
- National Register of Historic Places listings in Gilchrist County, Florida
- National Register of Historic Places listings in Lafayette County, Florida
- National Register of Historic Places listings in Cameron County, Pennsylvania
- National Register of Historic Places listings in Bennett County, South Dakota
- Incomplete list below:
- If a county has sublists (e.g. New York City), the script will only check the main county list shown in the table. To get around this, we can add new rows to these tables for the individual sub-county lists. For an example of this, see Cumberland County, Maine, in my sandbox. If we add a row for a sublist, the first cell (the "Code") needs to be "-----" (5 dashes), so that the mapping script ignores it. Also if there are duplicates across sublists in a county (also shown in the example in my sandbox), the Code for the duplicates row needs to be "ddddd" (5 d's). The script can account for these, and when it's totalling the entries up after validation, it knows to subtract duplicate rows and to only add the county total to the state total, not the individual sublists (which would be counting twice).
Also note that if you use the shortcut WP:NRHPPROGRESS to get to the Progress page, the script won't work. I'm about to ask at WP:VPT why that won't work because according to how Wikipedia's JavaScript variables are set up, it should. Not sure what's going on there. For now, just use the full title of the page, and everything should be fine.
Maybe if we update this list to workaround these limitations, the script can be reprogrammed to actually edit/update the page, but I don't want to do that until we know for sure that the script is outputting the correct data. Doing that now would erase a lot of the hard work several editors have put in, so that's still a little bit away but nonetheless still the end goal.
I ran it last night through this entire page, and it smoothly checked all 3000+ counties (and found more errors than I imagined btw, but I didn't correct any)... problem is it took more than an hour to do so. It validates at a rate of about 1 county every 1.5 seconds (for me.. I believe the limiting factor is internet speed, so it may be faster/slower for you). It's not lightning fast, but definitely an improvement over actually visiting all the county lists individually. I mean how long would it take to manually go through 3000+ county lists? Definitely more than an hour.
Because of the speed, I wouldn't suggest trying to validate the entire country at once like I did, but it is possible. I would suggest using it to validate individual states by opening up the edit window for that state and clicking the "Validate county data" button on the edit/preview page instead of on the full page.
The code is located at User:Dudemanfellabra/NRHPstats/sandbox.js. I've made it able to run side by side with the NRHPstats.js code, so there shouldn't be any fear of incompatibility; I have them both running now in my js. Eventually I think this code will replace NRHPstats, but since this new code is in the very early stages and NRHPstats has been tested to work without problems, I didn't want to replace it just yet. If you decide to use this code, please let me know if you spot any errors/limitations that aren't listed above, and I'll try to get them worked out. Thanks!--Dudemanfellabra (talk) 17:51, 11 April 2013 (UTC)
- After some help from WP:VPT, the script now works even if you use the shortcut to get to the Progress page. I've struck that out above.--Dudemanfellabra (talk) 18:39, 11 April 2013 (UTC)
- Wouldn't it make more sense to add zero-listings entries to the state lists rather than hard-code them into the script? Ntsimp (talk) 20:43, 11 April 2013 (UTC)
- Take a look at National Register of Historic Places listings in Alaska. There aren't any county sections on that page. Every single county/area has its own article except Wade Hampton Census Area, which doesn't have any listings. The state list would look weird if it just had that one section on it. The bot can easily be programmed to zero out all these counties. I just need their ID's and I can make an array to scan through. Just a few more lines of code. I think that's the easiest way to go.--Dudemanfellabra (talk) 21:00, 11 April 2013 (UTC)
- The only problem with that is if any of those counties get listings in the future, it breaks the code. It's fine if you're willing to check those counties every so often, but you'd have to account for it somehow. TheCatalyst31 Reaction•Creation 22:11, 11 April 2013 (UTC)
- Not necessarily. I could make the code only check if it's looking at one of those counties if an error is triggered. If something is listed in those counties later, an error won't be triggered, and thus the code wouldn't check to see if it's looking at that county. Even if that weren't true and we did have to manually keep track of these counties, there's only 59 counties total in the US that have no listings. Many of those 59 have sections on their state pages, so the number of "hard-coded" counties wouldn't be that large anyway. They could easily be checked manually. I still hold to the fact that I can make the code handle it, though.--Dudemanfellabra (talk) 22:29, 11 April 2013 (UTC)
- In fact, I just edited the code to correctly handle Wade Hampton, and I get no error when validating Alaska. Without changing the script, I then added a section for Wade Hampton to the Alaska state list to test it. I reran the script on the Alaska table and it performed as expected, telling me that what was entered into the table (0) didn't match up with what the script found (3). Then I removed the fake table and got no error again. This method can easily be extended to other counties by simply tacking on OR logic or cycling through an array as I suggested initially.--Dudemanfellabra (talk) 22:45, 11 April 2013 (UTC)
- The only problem with that is if any of those counties get listings in the future, it breaks the code. It's fine if you're willing to check those counties every so often, but you'd have to account for it somehow. TheCatalyst31 Reaction•Creation 22:11, 11 April 2013 (UTC)
- Take a look at National Register of Historic Places listings in Alaska. There aren't any county sections on that page. Every single county/area has its own article except Wade Hampton Census Area, which doesn't have any listings. The state list would look weird if it just had that one section on it. The bot can easily be programmed to zero out all these counties. I just need their ID's and I can make an array to scan through. Just a few more lines of code. I think that's the easiest way to go.--Dudemanfellabra (talk) 21:00, 11 April 2013 (UTC)
- Wouldn't it make more sense to add zero-listings entries to the state lists rather than hard-code them into the script? Ntsimp (talk) 20:43, 11 April 2013 (UTC)
- Looking at the NY differences, they are almost all .1% off, I didnt round while entering them, your script does. AFAIK, go ahead and update all of them. dm (talk) 04:01, 12 April 2013 (UTC)
Help?
I have over the past few days gone through states up to Illinois validating the data in these tables. When a county has sublists, I have been adding them according to the rules I laid out in the 4th bullet point above. My method has been as follows:
- Click the edit section link for the next state in line.
- When the preview page comes up, click the "Validate county data" button at the top of the list. The script will then go through each row and try to validate the data, showing you its progress along the way. After it is finished, the cells which don't match up with the script will be colored red. The script output is made available as a tooltip, i.e. scroll over the cell and the script output will be displayed.
- If a county has an error in the total number of listings, check first to make sure that the county does not have sublists. This is generally obvious if there are 100+ listings in a county and the script is telling you that only like 50 should be there. If the number is only off by 1 or 2, more than likely that county doesn't have sublists. If it does not have sublists, check the actual county list to make sure the script is not giving an error (it shouldn't be and never has, but I still check every time). If the script is correct, correct the row in the edit box then skip to step 7. If it does have sublists, continue to the next step.
- For each sublist, add a new row with first column "-----" and second column "$COUNTYNAME [[$SUBLISTARTICLE|$SUBLISTAREANAME]]" (e.g. "Cook: [[National Register of Historic Places listings in North Side Chicago|North Side Chicago]]"). You can use the NRHPstats.js script to fill in the data for these rows. If the main county page has listings on it that aren't split out anywhere else, make the $SUBLISTAREANAME "Other" (e.g. "Cook: [[National Register of Historic Places in Cook County, Illinois|Other]]"). Make this "Other" sublist come after all other sublists.
- Check the state list article to see if there are duplicates among the sublists of that county. If there are, add a row after the last sublist (usually the "Other" sublist), and give it the first column "ddddd" and second column "$COUNTYNAME: Duplicates", with a citation explaining which listings are duplicated, where they are duplicated, and if those listings are articled/illustrated. For examples, see the states already done.
- Add a "Total" for the county under any sublist/duplicate rows. The first column should be the "Code" that was there originally, i.e. the 5-digit number for that county.
- After all sublist rows have been added and all red cells taken care of, click the show preview button. Run the script again so it can check the totals for counties with sublists (and to make sure you didn't introduce errors by adding those rows) as well as the state total. If these are incorrect, correct them.
- Finally, show preview one more time and run the script again to make sure no cells turn red. If they do, repeat the last two steps until no errors are produced. Change the "Last updated" date at the top of the list; then save the page.
- After the page saves, update the "State totals" section at the top of the Progress page, making sure not only to update this state but also the national totals.
As I said, I've done this for all states through Illinois over the past 3 days, but I don't have as much time as I'd like to have to go through all the states. If anyone is willing to use the method described above to help out, it would be greatly appreciated. If not, I'll get it done eventually. After we add the sublist rows once, validation/updating should go really smoothly. I envision using the script to validate once every two months or so. As always, if you find something crazy in the code that produces an output of "error" or just breaks it entirely, let me know, and I'll see if I can fix it. Thanks!--Dudemanfellabra (talk) 01:07, 15 April 2013 (UTC)
- I started to do NY, could you double check I'm doing it right? Lots of broken out counties in that state due to the sheer number of places. dm (talk) 12:33, 16 April 2013 (UTC)
- If you include a link in the "Total" row, the script will try to look in that link, and it will give you an incorrect number. I just edited the counties you had split out to match up with other states, but there are still three more to split out in the rest of the state (Also, a few other counties had incorrect articled/illustrated totals that I didn't fix).--Dudemanfellabra (talk) 16:49, 16 April 2013 (UTC)
- Thanks! I'll go back and try to finish it now dm (talk) 02:10, 17 April 2013 (UTC)
- Ok, NY is done dm (talk) 03:59, 17 April 2013 (UTC)
- Thanks! I'll go back and try to finish it now dm (talk) 02:10, 17 April 2013 (UTC)
- If you include a link in the "Total" row, the script will try to look in that link, and it will give you an incorrect number. I just edited the counties you had split out to match up with other states, but there are still three more to split out in the rest of the state (Also, a few other counties had incorrect articled/illustrated totals that I didn't fix).--Dudemanfellabra (talk) 16:49, 16 April 2013 (UTC)
- I get many red cells in Utah, but it looks like the script is counting "Address restricted" as an illustration. Is that really what we want to do? Ntsimp (talk) 15:13, 16 April 2013 (UTC)
- I've been counting it for every state I've done. There was some talk about whether or not we should count that image but never really a consensus.. My opinion is since we allow lists to go on the fully illustrated list even if they have AR images, we should count them here.--Dudemanfellabra (talk) 16:29, 16 April 2013 (UTC)
- OK, so Utah's validated. That's going to cause a lot of color changes on the next map update. Ntsimp (talk) 17:34, 16 April 2013 (UTC)
- I've been counting it for every state I've done. There was some talk about whether or not we should count that image but never really a consensus.. My opinion is since we allow lists to go on the fully illustrated list even if they have AR images, we should count them here.--Dudemanfellabra (talk) 16:29, 16 April 2013 (UTC)
- How do we handle something like Wisconsin which has multiple tables on the state page? dm (talk) 03:18, 25 April 2013 (UTC)
- What do you mean? I don't see anything unique about Wisconsin.. the script can handle multiple tables on the state page. In fact, that's kind of the whole point of using the script.. if it couldn't handle those, it wouldn't be useful for like half of the country.--Dudemanfellabra (talk) 03:34, 25 April 2013 (UTC)
- Sorry - that was the first state I had seen like that. I reread your description above and saw that you already handle this. dm (talk) 04:18, 25 April 2013 (UTC)
- What do you mean? I don't see anything unique about Wisconsin.. the script can handle multiple tables on the state page. In fact, that's kind of the whole point of using the script.. if it couldn't handle those, it wouldn't be useful for like half of the country.--Dudemanfellabra (talk) 03:34, 25 April 2013 (UTC)
New map sometime?
I haven't been following the new update techniques in the section above. It looks complicated. Is there anyway to get a quick takeaway message?
Maybe we could update the maps, say aiming for May 1?
Smallbones(smalltalk) 04:19, 24 April 2013 (UTC)
- Not very complicated. There's now a second script. The first one just adds up the numbers we're claiming, but the second one checks against what's actually present in the county lists. Ntsimp (talk) 06:37, 24 April 2013 (UTC)
- Yea, sorry I've been busy for the past week or two, so I haven't been able to contribute much. I'm nearing the end of my semester classes, so all my professors are laying the homework on heavy. I have finals next week, so I'll be able to upkeep this more after then. I plan to go through all the states that haven't been updated to include sublist rows then. After all of the sublists are added, we can run the map script to update the map. Might be somewhere around mid-May before I can get all of that done, but it might be earlier. If anyone wants to help out, the section above details how to add sublist rows. After they're all added during this first run through, validating/updating should be pretty easy from here on out. Just run the script and update the red cells. Pretty sure it could be done at minimum once every two months, maybe even once a month or even more often.--Dudemanfellabra (talk) 07:16, 24 April 2013 (UTC)
First validation of all counties
I've just gone through and validated every county in the country using the validation script. I've come across several oddities which the script doesn't like:
- In Langlade County, Wisconsin, one of the 5 listings has two images shown in the "image" column. As far as I can tell, this is the only site in the country that does this. I feel like we should only have 1 image in each cell, but that's mostly because I can't figure out a way to code around this haha. It is possible to combine the two pictures into one image file for use in the county list.. I think that might satisfy everyone. Any input?
- Similarly, in Walworth County, Wisconsin (as User:Ntsimp pointed out) there is one listing which has an image in the "Summary" column. The image is of a historical marker that describes the site. Again I think this image should not be there. Transcribing the text makes much more sense to me.
There may be other errors, although I checked relatively thoroughly when I was going through. In the future, this validation should take far less time because all of the sublist rows are added in. That's what took most of the time during this first run through.
I'm about to update the maps now!--Dudemanfellabra (talk) 22:07, 3 May 2013 (UTC)
- Good work. Thanks much. I already removed that sign picture. Ntsimp (talk) 22:10, 3 May 2013 (UTC)
- Maps updated. Comparing the new maps to the old ones, even though the old ones are only about a month old, I can already see improvement in several states!--Dudemanfellabra (talk) 22:22, 3 May 2013 (UTC)
- Many thanks for all of the hard work. Now we all can see more clearly where to focus our attention! Altairisfar (talk) 22:46, 3 May 2013 (UTC)
- Maps updated. Comparing the new maps to the old ones, even though the old ones are only about a month old, I can already see improvement in several states!--Dudemanfellabra (talk) 22:22, 3 May 2013 (UTC)
Address Restricted sites
In my combing of these lists, I have found that some lists, like St. Louis County, Minnesota, and Pinal County, Arizona, have address restricted sites that do not incorporate the address restricted image. Should we add the image to all such sites (that do not already have an image)? Chris857 (talk) 02:45, 4 May 2013 (UTC)
- I know this has been controversial in the past. I added some in Wyoming and got reverted. And I have deliberately chosen not to put it on Emery County, Utah because the locations of the AR sites there have been published and there's only one left to get. Ntsimp (talk) 14:01, 4 May 2013 (UTC)
- Definitely, don't add the AR graphic by default. I'd say: leave it off unless and until you yourself have made a serious effort to come up with an illustration, and persuaded yourself that it's not going to happen. For me, at least, those blue, or at least not-fully-red, counties on the map are an incentive to search harder for sites or for representative images. I've removed all the AR graphics from Nebraska, just to make it clear to me where I need to expend more effort; other photographers may feel the same way about their home ranges. Ammodramus (talk) 20:26, 4 May 2013 (UTC)
Automatic updating
When this page was first created in late March/early April, there was a lot of talk about getting a bot to update the data here automatically. As time progressed, that goal of full automation moved closer and closer.
First there was User:Dudemanfellabra/NRHPstats.js, which automated the tallying of data from each county list. This sped up the process a great deal, as manually counting the sites was the most time-consuming part of the process. This script was limited, however, because one still had to actually visit each of the individual lists and manually add the data here.
Then came User:Dudemanfellabra/NRHPstats/sandbox.js, which took things one step further and allowed the user to skip visiting all of the county lists but rather show all the data on a single page. This sped up the process even more since we no longer had to wait for each page to load. This script–though better than the previous one–was still limited because the user still had to manually input the data on this page and save it.
Well now, I would like to introduce User:Dudemanfellabra/UpdateNRHPProgress (this is a link to the documentation; the .js file is linked within). This script (which is intended to replace the sandbox script above, but several people are using it, so I haven't deleted it yet) now fully automates the process of checking each list and adding the data here. Currently the script is configured to output the result to your sandbox, but I've tested it over the last few days and worked out most of the bugs, so I think it can go live (i.e. be configured to output the data to this page rather than a sandbox). I wanted to wait for others' input before doing so though. The great thing about the script is that anyone can use it, and it isn't tied to one machine like most bots. This means that if I get busy IRL or even decide to leave Wikipedia for good, the script will still work.
The new script generates a button at the top of this page, similar to the NRHPstats/sandbox.js script, which when clicked sets the script in progress. It takes on average about an hour and a half (for me... it is limited by internet speed) to run, as compared to the several weeks it initially took us to compile all the data manually. Theoretically the script could be run constantly, although I think maybe once per day would be the maximum desired. This script doesn't update the map (although there is another one that outputs the SVG data for that... and actually I'm planning on updating that one to be more automated as well), so that won't be updated every day. I think maybe once a month or once every two months is still reasonable for the map.
An example update can be seen with this diff in my sandbox, just run a while ago by me. If there are any questions, comments, or bugs found, let me know, and I will try to address them. If none are found, I'll modify the script to go live shortly!--Dudemanfellabra (talk) 03:03, 8 May 2013 (UTC)
- FYI, the map script has now been updated as well to output the result to Wikipedia:WikiProject National Register of Historic Places/Progress/SVG.--Dudemanfellabra (talk) 17:09, 8 May 2013 (UTC)
- I've just run the bot to edit this page for the first time rather than my sandbox. The diff consists mostly of adding spacing to the tables, but it did update some data. If no one has any objections, I will run it again tomorrow.--Dudemanfellabra (talk) 00:16, 9 May 2013 (UTC)
- Fantastic! Beautiful to see the graph updated. What's this talk about "leaving the project for good?" You can't do that until ... Smallbones(smalltalk) 01:12, 9 May 2013 (UTC)
- Haha the "leaving the project" bit was more of an extreme case (although I have left for several months before.. but that's a different story for a different day). My point was that for bots, if the "owner" (i.e. the one that coded it) goes MIA, the bot stops running. A good example is the AlexNewArtBot, which many projects (including ours) relied on to generate lists of new articles that may be under the project's scope. When User:Alex Bakharev suddenly went MIA a while back, the lists didn't update, and Wikipedia was without that service for several months before User:Tedder stepped in with his bot to fill the gap.
- In the early discussion concerning this page, several people mentioned the idea of a bot to update it every so often. The bot would require one person and one person only dedicated to running the script. With this script, however, I could suddenly go MIA like Alex Bakharev, and anyone that felt like it could still use it. All they would have to do is import the script into their personal javascript page, and they would become a new "owner" of the script. I feel like this is a much better method of updating, especially since I frequently get too busy to get on Wikipedia, especially during the school year.--Dudemanfellabra (talk) 04:04, 9 May 2013 (UTC)
- Lets hope it's not necessary, but I agree with your design decision about a single user dependency. I dont agree the map should only be updated once a month, perhaps weekly? Super impressed with your focus on this and how much it's inspired others to take pictures and create articles. dm (talk) 04:42, 9 May 2013 (UTC)
- Thank you for the awesome and much appreciated work you've put into this! Altairisfar (talk) 18:49, 9 May 2013 (UTC)
- I'd be interested to see May 10 versions of the maps, to see whether there is visible progress. Is Altairisfar pushing out the red zone further in Alabama? There were observable differences in the "articled" map between the April 8 and May 3 versions, in Alabama presumably by dint of Altairsfar pushing forward, and in California and Utah and some other western states by efforts of TheCatalyst31 and others, including me. Not sure where else progress can be discerned by visual inspection of the thumbnail pic sequence of the articled map or thumbnail sequence of the illustrated map (though there is obviously progress going on, as shown in the table changes). I'd like to be able to see if there is! :) Or just to see that the maps are pretty much the same. I do think this initiative is great--thanks Dudemanfellabra for doing this. --doncram 20:31, 10 May 2013 (UTC)
- Both maps just updated. I guess I'm fine with the idea of a weekly update. It looks like we're on an "every Friday" update scheme according to the past two updates. I can't promise to actually do it every Friday, but as with the update script, the map script is useable by anyone. If I don't do it, anyone else can...--Dudemanfellabra (talk) 00:07, 11 May 2013 (UTC)
- I'd be interested to see May 10 versions of the maps, to see whether there is visible progress. Is Altairisfar pushing out the red zone further in Alabama? There were observable differences in the "articled" map between the April 8 and May 3 versions, in Alabama presumably by dint of Altairsfar pushing forward, and in California and Utah and some other western states by efforts of TheCatalyst31 and others, including me. Not sure where else progress can be discerned by visual inspection of the thumbnail pic sequence of the articled map or thumbnail sequence of the illustrated map (though there is obviously progress going on, as shown in the table changes). I'd like to be able to see if there is! :) Or just to see that the maps are pretty much the same. I do think this initiative is great--thanks Dudemanfellabra for doing this. --doncram 20:31, 10 May 2013 (UTC)
- Thank you for the awesome and much appreciated work you've put into this! Altairisfar (talk) 18:49, 9 May 2013 (UTC)
- Lets hope it's not necessary, but I agree with your design decision about a single user dependency. I dont agree the map should only be updated once a month, perhaps weekly? Super impressed with your focus on this and how much it's inspired others to take pictures and create articles. dm (talk) 04:42, 9 May 2013 (UTC)
Since the new script can update the entire country automatically, I see no more need for the older script–User:Dudemanfellabra/NRHPstats/sandbox.js–which requires the user to manually update cells in each state table. As such, I'm planning on putting a {{db-user}} tag on it. From this search, it appears that User:TheCatalyst31, User:Ntsimp, and User:Dmadeo are still including this in their vector.js pages. I think these users will get notifications from this message, but if not, I will remove the script from their vector.js pages myself. I plan on leaving the NRHPstats script in place since it gives some nice information on individual county lists and may be useful in the future if new tables are added to this page or counties are further split up.--Dudemanfellabra (talk) 21:31, 21 May 2013 (UTC)
- I've updated my vector.js page, so that's one less person you have to worry about. Thanks again for doing all this, it's been incredibly helpful. TheCatalyst31 Reaction•Creation 21:45, 21 May 2013 (UTC)
- Ditto. Ntsimp (talk) 22:40, 21 May 2013 (UTC)
- Myself as well. dm (talk) 08:14, 27 May 2013 (UTC)
Near the start of this "Progress project" I noted that there is a database on Commons that gives similar info provided every day (almost) by the erfgoed bot which works with Wiki Loves Monuments. It currently gives a total of 48336 (54.47%) sites illustrated which is 32 more than our count (and 0.53% less). BTW click the sort button at [1] on "Country" twice to get "US" near the top to find things most easily. Then click on "US" to get the county list, also at toolserver.
I'm definitely NOT saying that our effort here has been duplicative wasted effort. For one thing we've checked erfgoed bot's assumptions and calculations. I think that the bot includes all entries on the pages, including delisted sites. Given the similarities between the two results, I'll suggest that we try to work with the erfgoed bot - if only for short term updates - and perhaps try to let other countries know how to do maps of our format. Note that User:Multichill originally programed the bot, but I think somebody else runs it now. I'll inform him of this conversation.
Until about May 5, our efforts seemed to have turned off the bot results for the US (no daily updating), but since then it has been working. Did we do anything on May 4? The page Wikipedia:WikiProject National Register of Historic Places/Unused images is still turned off and it might be useful to get that running again, certainly before WLM in September.
I have one suggestion for the bot to make it more useful to us. The toolserver US county lists is not sortable by state, but certainly should be. For example, one of the top entries is "Abbeville County, South Carolina" (in one field, not two). I suppose that the simplest, but ugly way around this would be to list it as "South Carolina, Abbeville County" but I'm sure there is a more elegant solution.
I hope to hear more on this subject.
Smallbones(smalltalk) 14:23, 10 May 2013 (UTC)
- We absolutely need to get Wikipedia:WikiProject National Register of Historic Places/Unused images back, since this is the main instrument of finding the new images for the lists and articles.--Ymblanter (talk) 17:29, 10 May 2013 (UTC)
- The Commons bot was run at 04:28, 10 May 2013 (I believe that is US central time, but I can't remember if I have commons set to show local time or UTC), and our script here was run last at 11:15, May 9, 2013 (and is running as I type this). It may be possible that in that ~15 hr time span that ~30 new images were uploaded. I saw that happen recently in Tennessee or Kentucky I believe.. about 100 new images were added in one run.
- What's probably more likely--though I don't know how the Commons bot is coded--is that the Commons bot is not configured to account for duplicates while our page here does. I would think there would be more than a deficit of 30 if that were the case, though. But then again we may still have some problems with duplicates here. In the past month (which is when most of the duplicate data was added) some articles may have been created/pictures been uploaded (or deleted) so that our numbers here are off a little. It's on the possible future improvements list of our script to be able to automatically tally these duplicates, but that's probably a long way off.
- Another possibility is the delisted sites, although I find that less likely, especially since their total number of sites is less than ours (by about 50). Maybe their total is pre-programmed and just hasn't been updated, whereas our total is dynamic. Maybe their bot looks at the individual articles instead of the lists, in which case some error is to be expected since some sites have images on the articles but not in the lists or vice-versa.
- TL;DR: Without knowing more about how the bot is programmed, it's anyone's guess as to why the numbers are different.--Dudemanfellabra (talk) 17:39, 10 May 2013 (UTC)
- I just ran the script for today's update. We added a total of 2 pictures (and 5 articles) since yesterday, bringing us up to 48,305 pictures total. So that rules out the ~30 pictures in one day idea.--Dudemanfellabra (talk) 20:02, 10 May 2013 (UTC)
- Hi guys, I'm still involved in the monuments database and Wiki Loves Monuments, but just as a normal volunteer. I haven't looked at it for quite some time and I now noticed some things are broken. I'm doing a manual run of the unused images now and see where it bugs out. Hope to get it fixed today. Everything I build is open so you can find the source of the bot in svn (source of the api too) and the database is public readable on the Toolserver (connect to p_erfgoed_p on sql) or you can download the daily dump. The database uses tables in the form "monuments_<country code>_(<language code>) to store all the sources. This is a one on one mapping for the templates. This information is aggregated in "monuments_all". The type is not aggregated so I have to do this query to find the types:
- I just ran the script for today's update. We added a total of 2 pictures (and 5 articles) since yesterday, bringing us up to 48,305 pictures total. So that rules out the ~30 pictures in one day idea.--Dudemanfellabra (talk) 20:02, 10 May 2013 (UTC)
mysql> SELECT type, COUNT(type) FROM `monuments_us_(en)` GROUP BY type; +---------------+-------------+ | type | COUNT(type) | +---------------+-------------+ | HD | 13481 | | NB | 8 | | NBP | 2 | | NHL | 1946 | | NHLD | 397 | | NHP | 19 | | NHS | 66 | | NMEM | 9 | | NMON | 28 | | NMP | 8 | | NRHP | 71452 | | NRHP-delisted | 1318 | +---------------+-------------+ 12 rows in set (2.21 sec)
- As you can see I have 1318 delisted sites in the underlying table. I modified the aggregation query to not include these delisted sites (
WHERE NOT `type`='NRHP-delisted';
). - Would be nice if you could use the data in the monuments database. It's harvested every night anyway so you don't have to do it yourself.
- What I would really like to see is these kind of maps for other countries too. What would you need to make a map like File:NRHP Illustrated Counties.svg say for example the Netherlands? Multichill (talk) 16:00, 11 May 2013 (UTC)
- As you can see I have 1318 delisted sites in the underlying table. I modified the aggregation query to not include these delisted sites (
- It doesn't help that I don't know the language you programmed this in haha, but from what I can extract from your above comment, it appears that you are counting/tallying up delisted sites as well as listed ones by simply finding all instances of the {{NRHP row}} template. Did I understand correctly that you just modified the bot to ignore delisted properties (i.e. ones with
|type=
? I ask because the commons page still shows the total number of sites including the delisted ones. Is this just because the bot hasn't run since your modification? If I add up the totals from your query above, excluding the delisted properties, I get 87,416 total monuments (as opposed to the 88,734 figure currently shown on the Commons page). Our number here gives a total of 87,784, which is 368 higher than your number. - Could it possibly be that 368 sites are not using the NRHP row template, and that's where the discrepancy arises? I know I recently edited one listing to use that template, so it may be possible that more don't use it. My script doesn't go into the wikitext to extract statistics but rather the parsed HTML code itself, so it doesn't matter if a site uses NRHP row or is manually added in with table syntax. Because of the way I've set up the code, I'm almost certain that the number it outputs is correct, so the only way I can think of to explain the discrepancy is the lack of NRHP row templates on 368 listings. Maybe we could somehow check? Or possibly run the bot you initially ran to try to find these sites?
- As far as the maps for other countries go: The maps here are generated according to the "Code" column on this page. The output for the SVG can be found here. The base map is commons:File:Usa_counties_large.svg, which is set up so that each county in the map is labelled by its FIPS code. I can make classes based on those codes and color each county. I use another script to extract the percentages from this page to generate the SVG output and then copy the output to the SVG file on my hard drive to upload it. I imagine there is probably a similar base map for other countries that can be used in this manner. For the Netherlands, maybe commons:File:Nederland_gemeenten_2009.svg would work? Looking at how the SVG is coded, it doesn't look very user friendly, but I'm sure it could be modified to be a bit easier.--Dudemanfellabra (talk) 18:36, 11 May 2013 (UTC)
- Dudemanfellabra, regarding Tennessee or Kentucky, it wasn't an error. About a month ago, I took a trip down the Mississippi River as far as northeastern Arkansas. Along the way, I picked up almost 20 sites in Tennessee and several dozen sites in Kentucky, and I took five counties in those states from virtually nothing to fully illustrated. Nyttend (talk) 19:41, 16 May 2013 (UTC)
- It doesn't help that I don't know the language you programmed this in haha, but from what I can extract from your above comment, it appears that you are counting/tallying up delisted sites as well as listed ones by simply finding all instances of the {{NRHP row}} template. Did I understand correctly that you just modified the bot to ignore delisted properties (i.e. ones with
Puerto Rico, Guam, etc.
Can the "illustrated" and "articled" maps please be updated to include status of Puerto Rico, Guam, Northern Mariana Islands, etc.? As shown in the "File:NRHP nominations upload status.svg" map, this appears to be technically possible, and the table here already does seem to be updated for their numbers articled and illustrated. Maybe this was gonna be done soon anyhow? I'd just like to be friendly to the Puerto Rico editors who are plugging along very well, but their progress is not shown. Note: each of these should just be posted at their aggregate performance, without any subdivisions. There are no counties defined in Puerto Rico; it consists of 78 Municipalities of Puerto Rico that do not require separate progress tracking; we use artificial divisions of "western" vs. "central" vs. "metropolitan San Juan" etc. in our NRHP list-articles; these artificial divisions should not be reflected in any progress map. So maybe this would be relatively easy? --doncram 11:41, 16 May 2013 (UTC)
- That's on my to-do list as far as the maps go. I would also like to include tables here so that the script can automatically update those percentages. I don't want to do the latter until the former is done, though. There is also the idea of adding an inset with the state percentages, suggested by User:Nyttend if I'm remembering correctly. The problem with adding things to the map is that I'm not using Inkscape or anything to make the map; I'm using a text editor (that's why the SVG output is in a text format).
- There are other maps (e.g. as you mentioned the Focus status map) with the territories and associated states included on them, but their nominal dimensions are different than the county maps here. Because of this, I would have to manually adjust all of the paths inside, and there's no doubt I would screw that up. I may download Inkscape or some other app and try to figure out how to save the file without messing up what I already have working. I'll try to work on that in the coming days, but I'll be out of town for this upcoming weekend, so even the script won't be run for those days unless someone wants to run it. As such, I'll be updating the map today instead of tomorrow as scheduled. Then next week the "every Friday" schedule should be back on.--Dudemanfellabra (talk) 17:12, 16 May 2013 (UTC)
- I found time earlier today to download Inkscape and play around with it. I added the territories that appear on the Focus map, although there are still a few other territories--e.g. Palau--that are tallied up but not included on these maps. I may see if I can add those later. I also added an inset to show overall state percentages. I added these percentages in manually to the map this time, but I will work on updating the map script to automatically update the states before next Friday.--Dudemanfellabra (talk) 20:44, 16 May 2013 (UTC)
- That is so cool! Great work as usual. Ntsimp (talk) 23:07, 16 May 2013 (UTC)
- Yes, a big improvement over what was already very good. It's particularly helpful when comparing states; going strictly by the county map makes me think that Wisconsin has more pictures than Illinois and that Nevada has tons more articles than states such as Illinois, so the additions are important to show the true state of things. Nyttend (talk) 00:41, 17 May 2013 (UTC)
- That is so cool! Great work as usual. Ntsimp (talk) 23:07, 16 May 2013 (UTC)
- I found time earlier today to download Inkscape and play around with it. I added the territories that appear on the Focus map, although there are still a few other territories--e.g. Palau--that are tallied up but not included on these maps. I may see if I can add those later. I also added an inset to show overall state percentages. I added these percentages in manually to the map this time, but I will work on updating the map script to automatically update the states before next Friday.--Dudemanfellabra (talk) 20:44, 16 May 2013 (UTC)
Thanks! And I've now added the subtotalling rows, and links, and so on, for automatically updating Puerto Rico, which covers 322 out of those 600 or so in PR, Guam, etc. I hope this is all you need for that, but i am not sure because i am not running the script myself. --doncram 03:54, 23 May 2013 (UTC)
- I've modified the script to account for Puerto Rico being automated. Because Florida, Puerto Rico has the section title "Florida", it was interfering with the state of Florida section on this page. To remedy this, I've added "listings" to all the state section headers here (e.g. it's "Florida listings" now instead of just "Florida"). This way the script can tell the difference between the state and the municipality. For the "Code" column, I've used "00000" and told the map script to ignore rows with that ID (it already ignores those with ID "-----"), so only the state total will be mapped.
- In the future if tables for Guam, Virgin Islands, Northern Mariana Islands, and American Samoa are added, they need to be added in the order that they appear in the state totals table (i.e. the next table that should be added is Guam, and the Virgin Islands should not be added until Guam is already there). These tables would need to use the 00000 ID like Puerto Rico does. Although the map does not include Federated States of Micronesia, Palau, Marshall Islands, Minor Outlying Islands, and Tangier (and it likely never will), the script can still automatically tally the data up if tables for those are added after the other 4 that are included on the map. The map script will simply ignore them.--Dudemanfellabra (talk) 00:27, 24 May 2013 (UTC)
Disambiguation links
The way my script is set up to run is that it initially assumes that every site is bluelinked in a list and then subtracts off the number of redlinks that it finds. This is because there is no special CSS class or other identifier for a bluelink, but there is a class="new" identifier on redlinks, so they're easy to pick out. This is how the script has been run every single time it's updated....... but I just realized a problem with that. What about disambiguation links? Surely if a list links to First Baptist Church, the script should not count that as "articled", right? Well, it hasn't been doing that, but thankfully dab links also have an identifier attached to them (class="disambiguation") which is easy to check for. As such, I just updated the script to also treat dab links as non-articles. I'm about to set the code running and go to bed (It's past 2 AM where I am), and it should finish in about 1.5-2 hours. The total article count will probably go down a bit, but I'm not sure how much. The diff the script generates can actually be used to find which county lists contain dab links that need to be fixed, so that's a plus... but the next map update will unfortunately probably go in the opposite direction than it has been going :\. There's also the chance that I just screwed everything up and when the script finishes, this page will be full of gibberish (though there's like a <1% chance that will happen). If that does happen, someone please revert the edit if I don't get to it first in the morning. Sorry for this obvious flaw in the code that has persisted through so many updates!--Dudemanfellabra (talk) 07:25, 27 May 2013 (UTC)
- Actually it seems as if I'm having some problems making the script do what I want it to do. When I ran the script, the articled count actually went up by 10 (partially because Pubdog created 6 articles in that short timespan), so it looks as if it still can't detect dab links. I did some testing and found out that I can find disambiguation links with the NRHPstats script, so the yellow box that shows up when that script is used on a county list will be correct, but if I use the UpdateNRHPProgress script, which loads county lists in the background, it can't find dab links. There must be something that Wikipedia does that "adds on" the dab class after the page is fully loaded that isn't happening when I load the page in the background. I'll do some more investigating there, but if anyone has any insight, please let me know. For now, though, the script will still run fine.. just bear in mind that the numbers are slightly inflated.--Dudemanfellabra (talk) 18:30, 27 May 2013 (UTC)
- Ok, I think I have everything sorted out. I've modified the code to be able to (hopefully) find dab links, but I'm not sure how much of a hit the speed is going to take. It will probably take longer than the usual 1.5 hours, but I couldn't even begin to guess how much longer. I'm about to run the code now, so don't be shocked if everything screws up haha. Also don't be shocked if absolutely nothing happens (again) because that just means there's an error in the code. I'll get this thing working sometime........--Dudemanfellabra (talk) 06:21, 29 May 2013 (UTC)
I've just run the updated code that subtracts out disambiguation links. I used the diff to correct links to 3 different dab pages and found one list to which an IP editor had tried to add a listing that was actually a contributing property and broke the formatting.
Much to my surprise, the addition of the code that checks for dab links has actually sped up the code, and quite dramatically.. the latest run was 42 minutes, compared to previous runs of 1.5-2 hours... Even though the code actually does more now (i.e. makes more queries), the code is overall faster. The only explanation I can come up with is that the "break" the processor gets when it stops to check for dab pages makes it not max out and thus the next page in sequence loads faster than if the break hadn't been there. Truly shocked me, but it is what it is. From now on, dab pages won't be counted as bluelinks.--Dudemanfellabra (talk) 09:13, 29 May 2013 (UTC)
Shortcut
I've found the article page so useful that I've created a shortcut for it List of RHPs progress I hope others find it useful. Smallbones(smalltalk) 03:58, 29 May 2013 (UTC)
- There's also WP:NRHPPROGRESS. Not sure if it's ok to have a redirect from main space to project space, but I'm not so well versed on policy...--Dudemanfellabra (talk) 04:06, 29 May 2013 (UTC)
DC first "state" to be fully illustrated
I just updated the maps for this week, and I would like to point out that as of the last update of the Progress page, DC is now fully illustrated. That makes it the first full "state" (using the term loosely) to be fully illustrated. Good work!
As an aside, North Dakota is only one article away from being the first state to be fully articled...--Dudemanfellabra (talk) 19:03, 31 May 2013 (UTC)
- Thanks to User:Farragutful for the last 2 pix (and many others). Barnstar on the way! Smallbones(smalltalk) 21:24, 31 May 2013 (UTC)
- I just moved Woodrow Wilson School (Fargo, North Dakota) into the mainspace, so now North Dakota is the first state to be fully articled too. Thanks to User:Doncram for writing the last article there, and many of the ones leading up to it. TheCatalyst31 Reaction•Creation 22:35, 31 May 2013 (UTC)
- Well, unfortunately for DC, the Capital Traction Company Car Barn was listed recently, and I don't believe we have a photo. Chris857 (talk) 23:40, 31 May 2013 (UTC)
- Actually, there appears to be a photo of the Capital Traction Company Car Barn, File:Decatur Street Car Barn.jpg. The article Capital Traction Company identifies this photo as that facility. I also used Google Maps street view and this portion of the building faces 14th St. NW near Decatur Street, NW. It's not terribly far from where I live so I can check it out next week (busy weekend). Farragutful (talk) 00:25, 1 June 2013 (UTC)
- Well, unfortunately for DC, the Capital Traction Company Car Barn was listed recently, and I don't believe we have a photo. Chris857 (talk) 23:40, 31 May 2013 (UTC)
- I just moved Woodrow Wilson School (Fargo, North Dakota) into the mainspace, so now North Dakota is the first state to be fully articled too. Thanks to User:Doncram for writing the last article there, and many of the ones leading up to it. TheCatalyst31 Reaction•Creation 22:35, 31 May 2013 (UTC)
The photo I mentioned above is part of the facility formerly named the Capital Traction Company Car Barn. However, it is at the far end of the building that appears to have a different address from the property listed on the NRHP. It is a rather large facility. Based on this information I wasn't sure what the designation included, so I took a picture of the part of the building where the appropriate address is located, File:Capital Traction Company Car Barn.JPG. Farragutful (talk) 00:54, 10 June 2013 (UTC)
- Well, it appears a new site was listed on May 1. I didn't do any fact checking, but in the latest update from the script, DC was no longer fully illustrated because of this site. It appears that since that edit, though, a picture has been added, so all is good once again.--Dudemanfellabra (talk) 03:09, 10 June 2013 (UTC)
This project inspired me to spend 12 hours driving around Chemung and Schuyler counties in NY -> two more fully illustrated
That's twenty more rural photos added. dm (talk) 05:46, 11 June 2013 (UTC)
- Congrats! I think it is still worthwhile updating Wikipedia:WikiProject National Register of Historic Places/full illus list
- Smallbones(smalltalk) 04:22, 13 June 2013 (UTC)
Anyone live in or near Youngstown, OH?
In the search of the Internet for pictures of NRHP sites, I found that there are several unillustrated sites in Youngstown (list) that don't appear to have free media available. It wouldn't be hard to get several of them; there are at least four unillustrated sites within about a block radius of this location. (Alternatively, if anyone knows of any sitting on Commons but not in the list or article; I have come across several like that) Chris857 (talk) 02:49, 21 June 2013 (UTC)
- Pretty sure there aren't any on Commons. I went through everything for Mahoning County last year in preparation for a photo stop in downtown Youngstown (it yielded eleven new pictures), but since I'm not in the area very often, I can't help in the near future. Still hoping to get more, since the northeast has more unillustrated sites than anywhere else in the state. Nyttend (talk) 21:40, 27 June 2013 (UTC)
Half-illustrated sites?
The weekly list of property changes for April 26 increased the boundary of the Buffington Island Battlefield, which was previously listed in Meigs County, Ohio to include part of Jackson County, West Virginia. Unfortunately, the only pictures we have of the battlefield are on the Ohio side, and considering that there's a rather large river between the two sides, it doesn't really make sense to have an Ohio picture in the West Virginia list. This causes some trouble for this page, as it's not clear whether its duplicate listing should be listed as "illustrated" or "unillustrated". Right now, I have it as "illustrated", but it's flagging the total as an error since I didn't subtract the nonexistent duplicate image and will probably break the total the next time the script runs. Listing the site as "unillustrated" would prevent the script from returning the wrong total, but it's also wrong, since the article itself has images. (Alternately, we could just stick an Ohio picture in the WV list anyway, but as I mentioned this has its own drawbacks.) How should this list handle these kinds of listings? TheCatalyst31 Reaction•Creation 06:15, 27 June 2013 (UTC)
- This same conversation occurred about a month ago between User:Nyttend and me in another part of the country. To be honest I don't think it was ever resolved, but I feel like we should have some kind of system to address this. The way to make the math work is to count it as unillustrated, but there could be some kind of explanation in the duplicate ref along the lines of "XXXX appears in List1 and List2, but the image associated with it does not represent List2 accurately and is thus not included there. For counting purposes, this site is listed as unillustrated." Obviously the wording can be changed, but something along those lines should suffice.
- Also, I've been busy for the past few days, so I haven't been able to run the script. I'll do that a little later today if I can.--Dudemanfellabra (talk) 18:03, 27 June 2013 (UTC)
- FYI, no increase was listed — all that happened, both for the Ohio portion and the West Virginia portion, was a determination of eligibility, so apparently the owner objected or some other hitch occurred. I agree with the idea behind "XXXX appears in List1...", thinking that we shouldn't use the same image on multiple lists unless the image shows portions on both sides of the boundary. For this reason, I think it's good to use File:County Bridge No. 45, western portal.jpg on two lists: it was taken from one county, but it shows the bridge all the way into the other county. On the other hand, the Shakamak State Park HD is on three county lists, and I've added separate photos for each of them, since the best one, File:Shakamak State Park gatehouse.jpg, shows a little portion that's only in one county. Nyttend (talk) 21:25, 27 June 2013 (UTC)
- In that case, I should probably take it off the Jackson County list and adjust everything else accordingly. TheCatalyst31 Reaction•Creation 00:18, 28 June 2013 (UTC)
- I have to differ with Nyttend on this—I'd use out-of-state/county photos in the list articles, although I'd seek a photo that's actually taken in the appropriate chunk of geography.
- Insisting that list-article illustrations come from the state or county of the list could give rise to complications. First, would that also preclude the use of a photo of a historical marker that's not on the historic site itself? I've illustrated several sites with such markers that are in a nearby town or on a nearby highway, especially when the site itself is inaccessible. Second, I'm seeking photos of artifacts to illustrate AR archaeological sites. If a site is split between counties or states, would this require me to establish exactly where an artifact had been found, and to use its photo only in that county/state list?
- I'd say: treat out-of-jurisdiction photos as we'd treat poor-quality photos—use them until we can find something better, and actively seek such better photos. Ammodramus (talk) 14:19, 28 June 2013 (UTC)
- Agree with Ammo, especially his last paragraph. I'll take this opportunity to say that almost all photos of markers and signs should also be considered similar to poor-quality photos - we want photos of the thing itself, rather than the sign that says "that thing that you didn't photograph is the important thing". Sometimes that can't be helped however, but generally we should try for a better photos. A recent example File:Field hospitals AdamsCo PA.JPG, it was getting dark and I had one more site I could get that day. Take a picture of the marker or the empty field behind it? A review of the nomination form later revealed that there is a church nearby that should be a good place to photograph for the HD. Smallbones(smalltalk) 15:54, 28 June 2013 (UTC)
- Artifacts are a different situation — with artifact photos, or photos of other things from the site (e.g. File:Steamboat Bertrand.JPG), you're depicting what the site is, not where it is; anyone will understand that a photo of a few projectile points wasn't taken in situ. Photos concentrating on historical markers aren't good; they aren't significant parts of the site and don't contribute to its significance, unlike artifacts, so I'd say that we shouldn't use them in the lists. All this is my opinion regarding markers on site; markers away from the site are more of a problem, since as markers are normally placed by the site that they commemorate (e.g. File:Greersburg Academy.jpg), and since they have a definite geographical context, someone will expect the site to include the marker, which obviously won't be the case in some situations. Nyttend (talk) 17:25, 28 June 2013 (UTC)
- Agree with Ammo, especially his last paragraph. I'll take this opportunity to say that almost all photos of markers and signs should also be considered similar to poor-quality photos - we want photos of the thing itself, rather than the sign that says "that thing that you didn't photograph is the important thing". Sometimes that can't be helped however, but generally we should try for a better photos. A recent example File:Field hospitals AdamsCo PA.JPG, it was getting dark and I had one more site I could get that day. Take a picture of the marker or the empty field behind it? A review of the nomination form later revealed that there is a church nearby that should be a good place to photograph for the HD. Smallbones(smalltalk) 15:54, 28 June 2013 (UTC)
- In that case, I should probably take it off the Jackson County list and adjust everything else accordingly. TheCatalyst31 Reaction•Creation 00:18, 28 June 2013 (UTC)
- FYI, no increase was listed — all that happened, both for the Ohio portion and the West Virginia portion, was a determination of eligibility, so apparently the owner objected or some other hitch occurred. I agree with the idea behind "XXXX appears in List1...", thinking that we shouldn't use the same image on multiple lists unless the image shows portions on both sides of the boundary. For this reason, I think it's good to use File:County Bridge No. 45, western portal.jpg on two lists: it was taken from one county, but it shows the bridge all the way into the other county. On the other hand, the Shakamak State Park HD is on three county lists, and I've added separate photos for each of them, since the best one, File:Shakamak State Park gatehouse.jpg, shows a little portion that's only in one county. Nyttend (talk) 21:25, 27 June 2013 (UTC)
50,000 illustrated sites by September 1?
Looking at our total illustrated sites - we're up to 49,243 (56%). A rough calculation shows me that we should be able to get up to 50,000 (~56.85%) in just about 2 months - just in time to start WLM-US.
Regarding WLM, we still have a problem with the upload buttons. To satisfy everybody, I suggest we put a "Photos needed" list on the talk page of every county (or similar) list and have the upload button there. Is there any way to automate this?
Also, I will not be coordinating WLM-US this year. I simply won't have the time. I do suggest that somebody from WP:NRHP does help out with the organization, as I felt that several of last year's organizers didn't understand the needs of WP:NRHP as well as they might have. In any case WLM-US in *not* exclusively a WP:NRHP project, but I feel that it should have some WP:NRHP-input. Smallbones(smalltalk) 14:21, 6 July 2013 (UTC)
drill down links within this page
I edited the State list section links to drill down to the state breakout tables, i.e. changing link from "National Register of Historic Places listings in STATE" to "#STATE listings". I think this is more helpful in general; i know this is more helpful for me personally, when using this page. I look at state level then want to drill down to county level. I think this should be changed also in the script, but looking at Dudemanfellabra's script i a) don't see where this is written out, and b) see that I don't have access to change the script. Could this possibly please be changed. Thanks. --doncram 19:41, 5 August 2013 (UTC)
Nebraska bug
There seems to be a bug related to the sorting of the table for Nebraska. When I try to sort the table by "% Start+", for instance, I get a string of 0%'s, followed by 1.0%,... in ascending order until I reach 26.7%; but then that's followed by 3.8%. I get similar results when I sort by different columns, both in ascending and descending order. It looks, more or less, like it's trying to sort by first digit. I've spot-checked half a dozen other states, and Nebraska seems to be the only one for which this is the case: the others sort in true numerical order. Unfortunately, I know exactly nothing about table syntax. I compared the initial code for Nebraska to that for Nevada, which sorts OK, but no differences jumped out at me. I'm reluctant to mess with the code, since I know so little about it. Could someone who understands table coding better than I look at this (and, if you can fix it, shoot me a suitable-for-a-rank-layman explanation of what you did)? Thanks. Ammodramus (talk) 00:55, 4 November 2013 (UTC)
- Using the ideas at Help:Sorting, I think I've fixed the problem. I forced all the columns to sort as numbers instead of as text strings (i.e. "2%" now comes before "20%"), and now it seems to work.--Dudemanfellabra (talk) 04:02, 4 November 2013 (UTC)
- Thanks for the fix! Judging by the number of characters you had to add, it took some work. I hope that you were able to automate it to some extent, and didn't have to spend a couple of hours hitting CTL-V over and over and over again.
- And once more, thanks for coming up with this page in the first place, and for all the work that you've put in on it. It's a great tool, and this editor for one is very grateful to you for it. Ammodramus (talk) 20:02, 4 November 2013 (UTC)
Separate pages for each state, then transcluded?
Would it make more sense to have these tables on separate pages for each state, then have them transcluded to this page? There is just so much information on this once page, and it takes a few seconds to load. Perhaps there are benefits to having all of the information on a single page. Also, are these tables updated manually or by a bot somehow? --Another Believer (Talk) 18:05, 8 November 2013 (UTC)
- The idea of having each state on its own subpage was tossed around when this page was being created--mainly as a way to focus discussion/cleanup/development on single states--but it never happened. I'm not necessarily opposed to the idea, but I would have to rework a lot of things to make that work. The page is currently updated by a script that I (and sometimes other editors) run every week or so, and the script is set up to only edit this page. I could rework it to output to individual pages, but I don't really see the benefit as far as loading time goes.. This page will still take the same amount of time to load (and perhaps even longer) if there are many transcluded subpages rather than one big page.--Dudemanfellabra (talk) 21:42, 8 November 2013 (UTC)
- No problem. Thanks for the explanation. I just wondered if it would be easier to fork the content and focus talk page discussions on specific states, but I am sure you have determined the best way to configure this project. I have to say, this is an incredibly impressive progress report. It would be very cool to one day see the maps in motion... the counties of the U.S. changing from blue to red over time. Keep up the great work! --Another Believer (Talk) 22:26, 8 November 2013 (UTC)
- You can do this by hand: just go to the Commons file for the map you're interested in, scroll down to "File history", and either download the old versions, giving them sequential filenames, or open the old versions in separate browser tabs and hit alt-tab repeatedly to go through them. Nice to watch the map progressively turning redder... Ammodramus (talk) 03:14, 9 November 2013 (UTC)
- That would be fun to watch, but it would take a long time to construct. If only there were a tool that could shortcut the process and create an animation like this... --Another Believer (Talk) 05:23, 9 November 2013 (UTC)
- There's almost certainly a way of doing that, but since the map is updated every week, we'd probably want to reserve making an animation like that for when we reach milestones. I'd definitely love to see a map like that when we hit 60% articled, for instance, or for the first year of having the progress page. (Unfortunately, none of the stats are especially close to reaching a round number, so we may have to wait a bit.) TheCatalyst31 Reaction•Creation 05:47, 9 November 2013 (UTC)
- That would be fun to watch, but it would take a long time to construct. If only there were a tool that could shortcut the process and create an animation like this... --Another Believer (Talk) 05:23, 9 November 2013 (UTC)
- You can do this by hand: just go to the Commons file for the map you're interested in, scroll down to "File history", and either download the old versions, giving them sequential filenames, or open the old versions in separate browser tabs and hit alt-tab repeatedly to go through them. Nice to watch the map progressively turning redder... Ammodramus (talk) 03:14, 9 November 2013 (UTC)
- No problem. Thanks for the explanation. I just wondered if it would be easier to fork the content and focus talk page discussions on specific states, but I am sure you have determined the best way to configure this project. I have to say, this is an incredibly impressive progress report. It would be very cool to one day see the maps in motion... the counties of the U.S. changing from blue to red over time. Keep up the great work! --Another Believer (Talk) 22:26, 8 November 2013 (UTC)
I was thinking about the animated map again recently. If anyone knows of a way to put the file history into motion, please let me know. It would be great to see the map progressively turning red over time, and it might even inspire others to turn their county red! ----Another Believer (Talk) 18:52, 26 July 2014 (UTC)
- DudemanF created an animated map a little while back. I forget where it is, but I remember that it was awesome! --Orlady (talk) 20:36, 26 July 2014 (UTC)
- All four are included at Wikipedia talk:WikiProject National Register of Historic Places/Archive 59#Animated Progress maps. TheCatalyst31 Reaction•Creation 21:47, 26 July 2014 (UTC)
- Oh, wow! How great! Thanks so much for putting this together and for pointing me in the right direction. ----Another Believer (Talk) 21:53, 26 July 2014 (UTC)
- Also see WP:NRHPPROGRESSHISTORY for other historical stats from this page (which I will hopefully update soon).--Dudemanfellabra (talk) 15:05, 27 July 2014 (UTC)
- Oh, wow! How great! Thanks so much for putting this together and for pointing me in the right direction. ----Another Believer (Talk) 21:53, 26 July 2014 (UTC)
- All four are included at Wikipedia talk:WikiProject National Register of Historic Places/Archive 59#Animated Progress maps. TheCatalyst31 Reaction•Creation 21:47, 26 July 2014 (UTC)
- DudemanF created an animated map a little while back. I forget where it is, but I remember that it was awesome! --Orlady (talk) 20:36, 26 July 2014 (UTC)
![](http://up.wiki.x.io/wikipedia/commons/thumb/1/12/NRHP_Counties_Net_Quality.svg/220px-NRHP_Counties_Net_Quality.svg.png)
I see the net quality map finally got made, and all I can say is, "say ya to da UP, eh!" I'm amused that the UP of Michigan is the only major region that is consistently red-colored. Chris857 (talk) 14:43, 22 November 2013 (UTC)
- Yes, I announced it at the project talk page shortly after adding it here. As to the UP, is that you working up there? I was wondering who created all those articles/took the pictures, but I never took the time to look it up. Yes, that area is consistently red in all maps of quality and quantity. If that was you, good job, sir!--Dudemanfellabra (talk) 18:58, 22 November 2013 (UTC)
- Yeah, I noticed it on the main project talk page after posting here, oh well. It isn't me, but rather Andrew Jameson mostly, who does a bang-up job on the whole state (but the UP is a lower-hanging fruit and is more complete). Because of that, even though I live there, I haven't been responsible for making it good. Chris857 (talk) 19:22, 22 November 2013 (UTC)
Untagged articles
I'm trying to check up on the untagged articles in Nebraska. According to the table, there are two such in Douglas County-- which has 116 articles. Is there a quick way for me to find out what those two untagged articles are, or do I need to go through and check talk pages one by one until I find them? Thanks-- Ammodramus (talk) 03:42, 25 November 2013 (UTC)
- If you have User:Dudemanfellabra/NRHPstats.js installed, a yellow box will appear at the top of the table on the county list outlining all of the stats listed here on the Progress page. The text for untagged/unassessed articles will appear in red. If you hover over that text, the tooltip will let you know which articles are untagged. Using that script, it appears First National Bank Building (Omaha, Nebraska) and Northern Natural Gas Building are the offending two.--Dudemanfellabra (talk) 03:57, 25 November 2013 (UTC)
- Thanks-- I don't have the package installed (if I'm even using the right phrase, which shows just how much I know about the subject). I'll take care of those two articles. Ammodramus (talk) 04:28, 25 November 2013 (UTC)
- If you would like to use it, simply add the following code to the bottom of User:Ammodramus/vector.js (which you'll need to create):
importScript('User:Dudemanfellabra/NRHPstats.js')
- If you don't want the yellow box to appear automatically, you can configure the script to add a link to the toolbox on the left side of the screen which you can click to bring up the yellow boxes. To do this, add
var NRHPstatsAuto = 'false'
- on the line before the importScript command. If you need more help with it, let me know on my talk page.--Dudemanfellabra (talk) 04:43, 25 November 2013 (UTC)
- Thanks-- I don't have the package installed (if I'm even using the right phrase, which shows just how much I know about the subject). I'll take care of those two articles. Ammodramus (talk) 04:28, 25 November 2013 (UTC)
Additional tables
After noticing that a NRIS-only article in Micronesia wasn't getting tagged by the bot, I added tables for the Federated States of Micronesia, Palau, the Marshall Islands, and the U.S. Minor Outlying Islands. There are now tables for every jurisdiction in the list except for Tangier (which is a unique case anyway). @Dudemanfellabra: you may want to check and see if any of the scripts need to be updated. TheCatalyst31 Reaction•Creation 06:33, 26 November 2013 (UTC)
- The new tables look great to me. A while back, I configured the script(s) to adjust automatically if new tables were added, but this is the first time there would be a "real" test of that. I'll run the script to update the data here some time before the end of the week to see if all goes well.--Dudemanfellabra (talk) 08:11, 26 November 2013 (UTC)
Georgia rural counties
I'm working on changing some of the rural Georgia counties "percent illustrated" from blue to red. Bubba73 You talkin' to me? 03:20, 13 March 2015 (UTC)
"Articled"?
"Articled" does not mean that something has an article. This is the English Wikipedia - let's use English, please. Bubba73 You talkin' to me? 06:30, 13 March 2015 (UTC)
- You're right that "articled" is not standard English, but bear in mind that this is not an encyclopedia page we're talking about. It's more like a scratch sheet that writers keep on their desks in the back office, but never intend to send to the printer for mass distribution. You'd expect to see informal language, in-speak, and abbreviated ways of saying things on such a scratch sheet. That's part of the purpose of the "Wikipedia:" namespace. On the other hand, it would be completely unacceptable to use "articled" on an encyclopedic content page. — Ipoellet (talk) 08:48, 13 March 2015 (UTC)
SVG original file not updated
The progress maps were recently updated. If you go to File:NRHP Illustrated Counties.svg and pick one of the sizes, the update is reflected. However, if you click on "original file", it is not updated. Is that the way it is supposed to be? I like to take the "original file", blow up a section of it, and print it out so I can see what counties need attention. If you pick one of the sizes and zoom in, the county lines get blury, but not with the "original file". Bubba73 You talkin' to me? 03:02, 14 August 2015 (UTC)
- If you pick one of the sizes, the image is rendered as a PNG file of a fixed size, while if you look at the original file it's rendered as an SVG file that automatically scales and resizes when you zoom in. The SVG file probably isn't updating because of a browser caching issue of some sort, so some combination of reloading and clearing your cache should fix it (depending on your browser); that happens to me from time to time after an update. TheCatalyst31 Reaction•Creation 03:25, 14 August 2015 (UTC)
- I brought it up, hot F5 to refresh, and that did it - thanks. Bubba73 You talkin' to me? 03:28, 14 August 2015 (UTC)
![](http://up.wiki.x.io/wikipedia/en/thumb/f/fb/Yes_check.svg/20px-Yes_check.svg.png)
Update maps?
I updated the statistics but the progress maps didn't update. How do you update the maps? Bubba73 You talkin' to me? 03:35, 1 September 2015 (UTC)
History?
The main page says "The History subpage shows historical maps and other data collected from this page." The maps all seem to be the current ones, i.e. not previous versions. Not that this matters much, but it could be an interesting comparison. Bubba73 You talkin' to me? 06:02, 2 September 2015 (UTC)
- The maps in the first section of Wikipedia:WikiProject National Register of Historic Places/Progress/History haven't been updated since March, but they are all animated maps showing how our data has changed over time. Does your browser do animated gifs? Ntsimp (talk) 11:20, 2 September 2015 (UTC)
- That works - I didn't know about that page. I was looking at the standard history of the progress page. Thanks. Bubba73 You talkin' to me? 16:33, 2 September 2015 (UTC)