Template talk:NRISref

note
This template is intended for use in all wikipedia articles having reference to NRHP NRIS database. It is being started to address need discussed currently at Wikipedia talk:WikiProject National Register of Historic Places/Archive 48. Help developing this template properly would be appreciated. --doncram (talk) 14:15, 5 November 2010 (UTC)
 * Testing :
 * and testing

parameter has been added for use in test cases. There are testcase here.&#32;– droll  &#91;chat&#93;  03:13, 3 July 2011 (UTC)
 * Looks good to me. I hate that this template is now protected. Maybe I'll look into becoming an admin. I'm tired of having to ask people to do things as simple as this.--Dudemanfellabra (talk) 15:42, 3 July 2011 (UTC)
 * ✅ &mdash; Martin (MSGJ · talk) 12:36, 4 July 2011 (UTC)

Sandbox update
The version in the sandbox updates the code to use tags so that the template page itself will not appear in the error category, Category:NRISref errors.--Dudemanfellabra (talk) 20:34, 1 April 2012 (UTC)
 * Yes check.svg Done Anomie⚔ 20:59, 2 April 2012 (UTC)

Grammatical error in the template
This template italicizes "National Register of Historic Places" (at, for example, Munson Diner). The National Register is no more supposed to be in italics than The Library of Congress, the Hollywood Walk of Fame or any other entity that is not a book, magazine, TV series or movie title &mdash; and it's certainly not italicized at National Register of Historic Places. As this is a protected article, would someone address this, please. --Tenebrae (talk) 00:07, 25 May 2012 (UTC)


 * As it's now been over a week and the previous edit requests were answered within a day or two, I'd like to request than an editor with privileges for this article please address this request as well. With thanks, --Tenebrae (talk) 16:11, 2 June 2012 (UTC)
 * To draw attention to your request, you should use the editprotected template. As for this specific request, the NRHP is being cited as a website and the convention in the "" style used by this template is that the title of the website as a whole (as opposed to the individual webpage) is displayed in italics. I'd want to see more discussion before trying to change that. Anomie⚔ 15:22, 5 June 2012 (UTC)


 * Of course. It does seem as if "" is an anomaly &mdash; article pages for websites themselves don't italicize the name (e.g. Salon (website), Deadline.com, Amazon.com, and countless others). Would this discussion take place at the "" talk page, or at some template noticeboard? --Tenebrae (talk) 23:37, 5 June 2012 (UTC)
 * Salon's article italicizes the name of the website. As for the others, they're also registered business names. The discussion page for the citation style is at Help talk:Citation Style 1 if you're still interested.  Imzadi 1979  →   12:06, 19 June 2012 (UTC)
 * Salon is an online magazine. No one italicizes The Library of Congress, the Hollywood Walk of Fame or the National Register of Historic Places, not even their Wikipedia articles. --Tenebrae (talk) 14:55, 16 February 2017 (UTC)

Edit request on 14 June 2012
Can someone add  to the template so that corporate authorship of the website by the staff of the NPS can be attributed? Thanks.

 Imzadi 1979  →   00:07, 14 June 2012 (UTC)
 * Code is on Template:NRISref/sandbox. I'm not sure that this is what you intend? &mdash; Martin (MSGJ · talk) 06:34, 14 June 2012 (UTC)
 * Yes, that's what I was looking for.  Imzadi 1979  →   07:06, 14 June 2012 (UTC)
 * I'm not sure that conveys any useful information? It is already attributed to the NPS ... I can add it if there is consensus for this change though. I've also got some more efficient code for the date formatting on the sandbox. &mdash; Martin (MSGJ · talk) 09:59, 14 June 2012 (UTC)
 * The publisher is attributed to the NPS, but as a staff-produced work, the authorship can/should be attributed to them as well.  Imzadi 1979  →   10:03, 14 June 2012 (UTC)
 * Personally I don't see the benefit, but I'll leave this request for someone else to deal with. &mdash; Martin (MSGJ · talk) 20:15, 15 June 2012 (UTC)
 * I agree with the above -- I'm not sure if this is necessary but have no objection to doing it. Just going to leave this request up for a while first for visibility/chance for objections/etc. Rjd0060 (talk) 02:42, 16 June 2012 (UTC)
 * No one else wanted to make the decision so I have declined this request as there does not appear to be consensus for it. &mdash; Martin (MSGJ · talk) 11:46, 19 June 2012 (UTC)

Option B: Please add: |author= to the template to at least allow articles that do include "Staff" as an author in other citations to include it in ones generated by this template for consistency.  Imzadi 1979  →   11:57, 19 June 2012 (UTC)
 * Why? Anonymous works do not get this sort of pseudo-attribution in any style guide I've ever seen, so it shouldn't be in this template's citation or other citations.  Nyttend (talk) 13:17, 25 June 2012 (UTC)
 * I agree with Nyttend.&#32;– droll  &#91;chat&#93;  21:26, 25 June 2012 (UTC)
 * Then I will no longer use this template in articles to avoid the inconsistency between attributing to the staff of MDOT or other agencies but not the staff of the NPS. Thank you.  Imzadi 1979  →   21:47, 25 June 2012 (UTC)

A simple way to cite a detail page in the NRHP?
The template {t{CRHP|17701|Nanticoke National Historic Site of Canada}} comes up nicely with links to the detail page for the site as See Template:CRHP Can we have an equivalent NRIS template that would display the details for an American site? It would be very useful on any site page, but particularly for pages like Reed and Stem which refer to many buildings. StarryGrandma (talk) 19:48, 26 August 2012 (UTC)
 * This sounds nice, but if I understand it correctly, I don't think the NRHP in DC works that way. So, e.g. if you put a registration number in a NRISref template - some of them might be able to go to the clunky NRHP focus site and get some pages (though I doubt it would be smooth) and others might have to go to to 25 different state SHPO sites.  I would guess that the Canadian project has a well organized web-site that they can connect to.  Smallbones( smalltalk ) 02:44, 27 August 2012 (UTC)


 * I like how the CRHP template provides two clickable links, one to a specific Canadian external URL that provides detail for the specific site, and the other to a wikipedia article Canadian Register of Historic Places about the their system. I wonder if we could at least have NRISref link to a new wikipedia article National Register Information System (surprised to see this is currently a redlink).  Which article could explain clearly that NRIS is a database, that it is downloadable, etc., using http://nrhp.focus.nps.gov/natreg/docs/All_Data.html, NPS's page on NRIS, as one source.  And goes on to discuss how NRIS data is used in many public webpage systems such as Archiplanet and www.nationalregisterofhistoricplaces.com and wikipedia.  The current NRISref has just one clickable link which goes from "National Register Information System" to that NPS page.  It's not a good quality NPS page for general readers.  It's not a good destination for 40,000 or so current NRISref appearances in wikipedia articles.
 * Thanks StarryGrandma for making the suggestion here. I hope we can improve NRISref template now to be more comparable to the CRHP template. -- do  ncr  am  15:50, 28 August 2012 (UTC)
 * You should certainly get rid of the redlink. Smallbones( smalltalk ) 21:13, 28 August 2012 (UTC)
 * There's no way to link directly on the focus detail page because the page is created dynmaically directly from the data base, it even does not have a perma link. The only linkable pages are PDFs of the nomination forms and accompanying photos what makes sense only if they're digitalized already and available, of course.
 * Maybe one should request such a feature from the NPS but I guess that they won't have any funding for that and at least as long as the Republican dominated congress exists that won't change. --Matthiasb (talk) 15:28, 29 August 2012 (UTC)
 * Yes, I agree that there is no exact equivalent to the nice CRHP link to a permanent webpage for each Canadian site. It's not possible to fix up this NRISref template that way.  And we do need to keep the NRISref template for purpose of giving reference to the NRIS database, when it has been used as a source.  Maybe there are two things possible:
 * Improve the existing NRISref template so that it serves readers better (does not direct to a confusing, non-informative NPS page)
 * Develop a different NRHPlink or other template that has a state-specific parameter, that accepts an NRIS refnum, and that works to link to permalinks for sites in states where permalinks are available. Which could be used in the way StarryGrandma wishes, to quickly and easily provide a link to more information on an NRHP, within a list of NRHPs in an architect article or elsewhere.  Maybe this could be set up and be functional for a few states.  To make it work for some states' sites, it would require a table of correspondence, and lookup function, to look up what is the state number used in its permalink, if that is different from the refnum. -- do  ncr  am  16:03, 29 August 2012 (UTC)

Have I got this right? - the template just generates a standard link to a page that describes what the NRHP database comprises, but doesn't actually give a link to the historic place itself? Seems pretty useless to me. If the various databases hold details of the places, however limited, it might be an option to convert those into a template. Otherwise it's not much of a citation, and I daresay other readers will be as fazed as me at its uselessness. John of Cromer in Philippines (talk) mytime= Mon 18:07, wikitime=  10:07, 11 March 2013 (UTC)


 * Yes you got that right. See next section for a small improvement. --06:05, 21 March 2013 (UTC)

Improvement to standard NRISref
Per discussions above and an ongoing one at Wikipedia talk:WikiProject National Register of Historic Places, there is some general recognition that the current standard NRISref is unsatisfactory. Relatedly, I started Wikipedia article National Register Information System, with intent that it should be linked within an improved NRISref, and that article has now survived an AFD.

So now, here i am to suggest that the current reference be changed to link to that article, where it now links to an external webpage where the entire NRIS database can be downloaded (but no readers would want to do that, and they should not be steered toward doing so).

The Wikipedia article can and should include some guidance / links about using an available search query form. And it can and should include guidance about using the www.nationalregisterofhistoricplaces.com online private copy of NRIS. The proposal here is to change the NRISref now to display an internal link to that Wikipedia article, right where an external link to this external page appears currently. -- do ncr  am  06:05, 21 March 2013 (UTC)


 * Still doesn't sound particularly satisfactory - what is needed is a direct link to the property whose reference number we already have. No need for counties or anything.  And hey, 4000 queries per year!


 * John of Cromer in Philippines (talk) mytime= Thu 14:56, wikitime=  06:56, 21 March 2013 (UTC)


 * Well, it's not possible to link directly to search results specifically about a given property. It would be an option to link to nothing, i.e. to mention "National Register Information System" with no link.  Would you prefer that, instead?  I would somewhat prefer to link to National Register Information System instead, and also to revise that article to serve this purpose better.  Yes, the 4,000-queries mention there is lame.  It was included by me to help the article survive the AFD, which is different than serving readers. -- do  ncr  am  07:03, 21 March 2013 (UTC)

IMO, the current version of the reference citation is still appropriate. As a reference, NRIS is much like a printed book -- it's not possible to link directly to the source of the information, but the reference citation should provide the information a person would need in order to get the information. The current citation links to the NPS page that gives the links for downloading the database; that seems like the best way to present the citation. A link to a Wikipedia article should not replace the link to the "publisher", IMO. It would be good if this could be done the way it's done for a newspaper/magazine article, wherein the reference citation links the article title to an online edition of the article and links the publication name to the Wikipedia article about the publication. In this instance, it seems to me that "National Register Information System" needs to link to the external website, but "National Register of Historic Places" could link to the National Register article on Wikipedia. --Orlady (talk) 23:36, 22 March 2013 (UTC)


 * FWIW, there is NRHP Focus that can link directly to the search results for a given reference number. It's not ideal but it is a step closer than just linking to the download link. Maybe that template can be integrated here for when a reference number is given? We could make a parameter refnum here that accepts the reference number. If it is set, link to the search results; if not, just do the standard citation. For example,
 * would yield:
 * and  would yield:
 * This could also be modified to add a name parameter like Template:NRHP Focus uses now which would make the template show the property name rather than the reference number.. whichever works best.--Dudemanfellabra (talk) 19:23, 23 March 2013 (UTC)
 * This could also be modified to add a name parameter like Template:NRHP Focus uses now which would make the template show the property name rather than the reference number.. whichever works best.--Dudemanfellabra (talk) 19:23, 23 March 2013 (UTC)
 * This could also be modified to add a name parameter like Template:NRHP Focus uses now which would make the template show the property name rather than the reference number.. whichever works best.--Dudemanfellabra (talk) 19:23, 23 March 2013 (UTC)


 * The first option with search results is bad because very often it links to crappy NPS page that just states something like "document unavailable, has not been scanned". Also that is NOT the source actually used by almost all nrhp editors and it does NOT show more than a small fraction of the nris data available and used for the wikipedia article. So it just looks like a mistake that it was used, when it is examined.
 * The second option is bad because it just links again to the NPS page where the nris database could be downloaded, which no reader should ever do, and it is silly to be suggesting it to them. --Doncram (talk) 20:57, 2 June 2019 (UTC)

Automatic link to weekly listings
I've got some code at User:Dudemanfellabra/Test which will generate a link to the weekly list page corresponding to a given listing date. It works with sites listed back to 1983 and will produce blank for dates before that. Here are some examples: Does anyone object to me adding the ability to show the link to the weekly list into the code? I think another good addition would be the link to the Focus search output, discussed in the section above.--Dudemanfellabra (talk) 00:01, 24 October 2013 (UTC)
 * 1) October 1, 1955 (or any pre 1983 date) -
 * 2) January 1, 1984 -
 * 3) December 31, 1995 -
 * 4) January 1, 1996 -
 * 5) September 11, 1999 -
 * 6) April 5, 2006 -

Update for new focus
Now that focus returns meta data for newer listings, can we add support for dates > July 8 2010? It would save having to look for the weekly list to add a reference by hand. A generic reference to the new focus should do. Generic1139 (talk) 02:24, 17 November 2015 (UTC)
 * The bottom of all Focus searches show "Last updated: 11/2/2013", so I've added that as version=2013 or version=2013a. I've also added parameters for adding the refnum and/or name of the site in question. If you add a refnum, the link will be changed to the specific Focus result in question.--Dudemanfellabra (talk) 00:07, 18 November 2015 (UTC)
 * Thanks. The common footer on the focus search pages does indeed say 11/2/2013, but it seems to contains listings through 9/30/14. --Generic1139  15:33, 19 November 2015

Here about the same. How is that supposed to work? What you're talking about is currently an undocumented feature. To refer to this Focus page (about Daniels Recital Hall, 2011 NRHP-listed in Seattle WA), I am here trying: which yields:

One way this seems odd is that the reference starts out "Staff (2013-11-02)", while I don't see any "Last updated: 11/2/2013" on the Focus page now...I suppose the Focus page display has changed since November. that does appear after all, though who knows what it refers to. It's a version of what? Obviously the NRIS database has been updated since that date.

My other alternative is now template:NRHP Focus? Which doesn't give documentation/instructions. Guessing, I'll try this:

That turns out better, it displays only "Staff. "NPS Focus: 10001105". National Register of Historic Places. National Park Service." with two appropriate links. That's what I'm going to use for now, and I think this is what should be advised at the NRISref documentation page here. Side question: on the Focus page there is, for this site, "Date Published: 1/3/2011" which I take to mean it was NRHP-listed on that date. Is that correct? How would it show differently for publishing of an owner-objection or other non-listing (or maybe those would simply not be posted at Focus, though they are in NRIS)? -- do ncr  am  01:51, 17 June 2016 (UTC)
 * If you add a refnum to the template, the link is generated by the NRHP Focus template, which points to the Focus result page, rather than to the generic download center (see below for more discussion about that). Additionally the refnum is also displayed in the link text. You can also provide a name for the site which will display before the refnum. I'll add this to the documentation soon.


 * As for the dates, this template was created when NRIS was still a big database that could be downloaded and was only updated every few years. In the past year or two it seems the NPS has begun to update the database more incrementally, and so having "versions" seems to be an obsolete feature actually. Maybe we should just get rid of that feature all together and not display any date at all?


 * I will post at WT:NRHP about the state of the Focus database overall to attract more attention, so please comment there for issues not related to this template specifically.--Dudemanfellabra (talk) 09:48, 17 June 2016 (UTC)

Suppress categorization
Is there a way to suppress the appearance of the template's testcases page Template:NRISref/testcases in the "Category:NRISref errors" maintenance listing? Fortguy (talk) 07:02, 7 October 2016 (UTC)

Authorship vs. publisher in format
I've only learned about this template today while reviewing the content of an article that uses it. I was surprised to see that "National Park Service" is apparently listed as an author, even though the cite format properly indicates it also as the publisher of the information. Surely there must be a reason why the organization is given the designation of author when clearly some particular person, unknown, collected and reported the information. I'm not sure authorship is needed at all — and especially since the NPS is indicated as the responsible entity for the information. I suggest removing the redundancy, eliminating the leading author information, and moving the wikilink to where NPS is listed as the publisher. --SidP (talk) 21:14, 13 September 2017 (UTC)

Which database version
How does one know which version of the database to include? Should any new articles use the most recent date? Leschnei (talk) 12:42, 25 September 2017 (UTC)
 * Plus one on this question. I think it's a bit unclear. Do you use the "date published" year on the NRIS article? I see no "database" phrase on a NRIS article. Example: --Engineerchange (talk) 05:00, 17 June 2021 (UTC)

add wikilinks
(Removed pending discussion below, I didn't anticipate this needing a discussion.)

Please add links to National Register of Historic Places and National Park Service in the work and publisher field s of cite web. This is supported by MOS:DUPLINK where it states: a link may be repeated in ... footnotes
 * This requires changing:


 * to:

Thanks, - Paul T [//en.wikipedia.org/w/index.php?title=User_talk:Psantora&action=edit +]/C 13:35, 17 April 2018 (UTC)
 * What MOS:DUPLINK is trying to say is that a link can be repeated in separate footnotes, or repeat in footnote if linked in the body. Doesn't mean a link should be repeated twice in one footnote. Galobtter (pingó mió) 14:07, 17 April 2018 (UTC)
 * I don't see National Register of Historic Places linked in this template, is it? I see that National Park Service is normally linked, but is sometimes not if no. Regardless, I will amend the request to remove this duplicate link. Is this better ? - Paul T [//en.wikipedia.org/w/index.php?title=User_talk:Psantora&action=edit +]/C 16:27, 17 April 2018 (UTC)
 * Yes, maybe make the national register of historic places also not linked if no Galobtter (pingó mió) 16:31, 17 April 2018 (UTC)
 * Nevermind, apparently the option is only used 1 out 60000 times..so it doesn't matter. I've ✅ the change and removed the unused options too.. Galobtter (pingó mió) 16:59, 17 April 2018 (UTC)
 * Thanks. I've restored my original request so you don't look nutty. Best, - Paul T [//en.wikipedia.org/w/index.php?title=User_talk:Psantora&action=edit +]/C 18:19, 17 April 2018 (UTC)
 * I've restored the option to use no. In many articles, wikilinks are not repeated at all in the footnotes, yet removing that parameter removed the ability to suppress a wikilink if this template was used a second or third time. This has become fairly standard, see the google maps citation template, among others.  Imzadi 1979  →   00:40, 18 April 2018 (UTC)

Link to nrhp.focus.nps.gov appears to be broken
From what I can tell, in certain circumstances, the template generates a link to the URL http://nrhp.focus.nps.gov/natreg/docs/All_Data.html. It appears that this URL is not functional. For determining a possible alternative, this NPS page may be of interest. --Elegie (talk) 05:51, 18 July 2019 (UTC)

The article references this url, http://nrhp.focus.nps.gov/natreg/docs/All_Data.html. It appears dead for me. --Engineerchange (talk) 05:12, 17 June 2021 (UTC)
 * Also, I notice that there is a broken template call at the top of the page. Not sure if that's expected behaviour. --Engineerchange (talk) 13:55, 17 June 2021 (UTC)
 * As far as I can tell, User:Jonesey95 fixed this issue in January of last year. Are you still seeing the  link somewhere? As to the "broken template call at the top of the page", that's just what happens when you view a template that requires a certain parameter but doesn't provide a default. It's quite common.  -- Tamzin (she/they) &#124; o toki tawa mi. 23:35, 18 June 2021 (UTC)
 * Adjacent to 2009a version under the Parameters section, I see the above link http://nrhp.focus.nps.gov/natreg/docs/All_Data.html. When I click it, I go to a "DNS_PROBE_FINISHED_NXDOMAIN" Chrome error page. --Engineerchange (talk) 00:37, 19 June 2021 (UTC)
 * Oh, I see what you're talking about. That's just the template's documentation page, not the template itself. For future reference, anyone can edit those pages; you don't need to put in an edit request. See Template documentation for more information. I've gone ahead and edited Template:NRISref/doc to note that the link in question is dead. I'm not removing the mention of the link, since knowing about it may still be useful to some people. -- Tamzin (she/they) &#124; o toki tawa mi. 00:45, 19 June 2021 (UTC)
 * Excellent, thank you for the response. I should've known the docs were for all to edit. Cheers! --Engineerchange (talk) 00:53, 19 June 2021 (UTC)

date format (2019)
This template has dateform which is optional. If not used, the default format for the date is year-month-day. Since this is a US-centric template, and most NRHP article use month-day-year format in dates, this template should default to mdy.

It is common for editors to use mdy format when writing NRHP articles, and forget to specify the format in this template (the Elkman lets it default) which leads to inconsistent date formats within the article (see MOS:DATEUNIFY). This particularly annoys one editor who winds up "fixing" the inconsistent format for every NRHP article that gets to DYK. MB 15:30, 8 September 2019 (UTC)
 * I agree whole-heartedly with MB. Could an administrator please change the default?  I also see the default formatting as jarringly different than the MDY formatting which is used by me and most other NRHP article editors (who are mostly U.S.-based, too).  Notably the MDY format is the default in references to NRHP nomination documents, as prepared by the "NRHP infobox generator" (see wp:NRHPhelp for more info if you want) used by me and many other editors in creating new NRHP articles.  I just inserted a edit request to this section, to ask someone to help out here.  I think this change would not be controversial. --Doncram (talk) 00:27, 22 September 2019 (UTC)   P.S. The request should have been by, and was changed to, edit template-protected.  Thanks Spintendo! --Doncram (talk) 02:35, 22 September 2019 (UTC)
 * , can you change this. Thanks. MB 00:46, 22 September 2019 (UTC)
 * I have made some changes to the template sandbox. Here is a diff showing the changes. I removed the display of a fake "January 1, 1900" date, leaving a blank date empty instead. I set MDY as the default date format. Please see Template:NRISref/testcases for the differences between the live template and the sandbox version, and feel free to add useful testcases. If these changes look good, I will move them to the live template. – Jonesey95 (talk) 04:12, 22 September 2019 (UTC)


 * Thanks Jonesey95! Great about the date format change, I suppose (though I haven't really looked at the testcases). About your change implementing a new policy "Do not display date if no date is provided", I actually have a problem with that.  Sorry this posting is a bit long, i am just trying to be precise and clear.


 * A version date/code should be provided, and I think if it is not provided (or if it doesn't match one of the dates/codes for one of the few known downloads) then something jarring should be displayed, and the reference/article should be put into an error category. Despite some confusion in some sections above, the NRIS database is not available online in any live, continually updated fashion which might reflect new listings and reflect corrections to old info (there exists tons of bad info in the database, see wp:NRIS info errors for documentation of thousands of errors, only some of which have been corrected in the offline current database that exists at the National Park Service).  NRIS is only in fact available to the public in relatively few discrete downloads which each were created by a NPS staffperson on a specific date (on just a few occasions) and each reflects all the good and bad information implemented into the NPS's private off-line database as of that date.  It is useful to indicate the provenance of the info in articles.  A more recent version date for an old listing indicates a little bit higher likelihood of the info being correct, because (i hope) only corrections are made and new errors are not introduced into old listings (i hope).


 * Most new NRHP articles are for sites listed before the end of 2014 in fact, and these come in with semi-automatically generated NRHP infoboxes, from editors using the "Elkman" NRHP infobox generator system, with proper identification of the database version/date from which they are generated (with coding NRIS2013a). Currently that system works for NRHP listings up to I think 9/30/2014.  All new articles for more recent listings are created without use of NRIS as a source, and therefore NRISref should not be used at all in their articles.  If NRISref is used, something should blow up.  Showing a more specific error would be better, though, than showing the default 2010-07-09 date.


 * One or a few editors recently using NRISref without specifying a date, on a bunch of new articles for recently listed NRHP places, has produced some inappropriate-looking articles, where the NRHP listing date is in 2017 or 2018 or 2019, but the reference supporting that shows date 2010-07-09. This version of an article for Jonesboro U.S. Post Office and Courthouse, is an example.  Obviously a 2010 source cannot support a 2019 event.  The editor in fact did not use NRIS as a source.  For the NRHP listing date, the refnum, the address, and maybe just a tiny bit more info about the property, they were in fact using the corresponding weekly NRHP listings announcement page at the NPS (e.g., this weekly announcements of May 31, 2019 page for that Jonesboro post office.  Those pieces of info will eventually show up (perhaps with additions of error, and definitely with many more fields of info) in a future downloadable version of the NRIS database. But, when NRISref was used in the article, the current coding which displayed the default 2010-07-09 date did a good thing, actually, in effectively reporting an error.  The solution going forward in that case, to which the editor of that article agreed, was for them to start to use a reference directed to the NPS's weekly listings announcements instead.  Which is facilitated by template:NRHP weekly, instead.


 * The default 2010-07-09 date is probably correct almost always where it does appear, which is probably in some number (very few? or thousands?) of articles about NRHPs listed before that date, probably all in articles created using the NRHP infobox generator before its underlying version of NRIS was updated to a later date (sometime in 2011 i think, and definitely by 2012). The exception is a few recently created articles for sites actually NRHP-listed at dates after 2014-09-30 (my guess is fewer than 100), in articles created by just one or a few editors.  Those exceptions should be found and fixed.  I wonder if AWB could be used to find those?  Or should a bot be requested to detect them?  It was sensible at the time that NRISref was started for that 2010 date to be the default.  Maybe now the date should be explicitly put into those old thousands of articles' usages of NRISref, which could be done by a bot.  Or could be done by an editor using AWB, if those pages were now put into an error category by a new change to NRISref.  I guess I would be willing to do that myself, though maybe not happily if there were too many.  To detect how many cases that would be, the NRISref code could be temporarily changed to put all those instances into an error category, say Category:NRIS version not specified (currently a redlink) and then we could see how many those are.
 * For convenience of an AWB editor, it might be more elegant/helpful if that could be split into one error category for old articles created before 2012 (or with NRHP listing dates before 2010-07-09) vs. one for newer articles where NRIS was really not used, but I don't see how that split could be implemented just by coding within NRISref, it seems to me that a bot run would be necessary to make that split. Or the AWB editor could just deal with having a mix of types of cases to fix. --Doncram (talk) 06:20, 22 September 2019 (UTC)
 * Please do look at the testcases page that I linked above. Here are some responses to specific items above, for the sandbox version:
 * I'm pretty sure that there is no default date in the sandbox version. The current version has a nonsensical default date of "1900-01-01" if the parameter is invalid or left blank.
 * If the date is left blank in the sandbox version, no date is displayed.
 * If the date is left blank or invalid, a red error message is shown, with a link to the help documentation. I added this just now.
 * When the version date is wrong or omitted, Category:NRISref errors is placed on the affected page, which already happens in the live version of the template. There are no articles currently in that category.
 * I see that you have an issue with editors using a version that does not make sense based on context. Perhaps after this basic formatting change is done, someone could come up with a way to apply logic to this template to check for that sort of error. I think it is out of scope for this basic formatting change, however, and you risk derailing this relatively simple improvement if you pursue it in this talk page section. I recommend that you start a new section on this page, with some concise, coherent thoughts and recommendations. – Jonesey95 (talk) 17:17, 22 September 2019 (UTC)
 * I agree the scope of this change was just the default date format. Everything looks good to me, I think you should make this live. One minor related thing is that the Category:NRISref errors isn't restricted to articles - could you fix that like you did elsewhere. MB 20:35, 22 September 2019 (UTC)
 * I have updated the template from the sandbox, including limiting the error-checking category to article space. I encourage to start a new discussion about any additional changes that are desired.  – Jonesey95 (talk) 22:34, 22 September 2019 (UTC)

towards improving NRISref reference
There's been lots of well-deserved criticism about this NRISref reference, expressed in numerous places. One huge issue is that the current version unhelpfully directs readers to a National Park Service webpage where someone might download a copy of the NRIS database, which is rude and/or mean and/or (insert other negative term here) to do. This template is protected for editing by administrators only, too, so I can't just fix it up better myself. And it has always been complicated to try to fix anything here, in part due to persistent misunderstandings about the nature of sources and the nature of databases. (E.g. some editors wish for any reference to be replaced by a link to info specific about a given NRHP-listed property, which cannot be done because you cannot link to a specific line in the database, and linking to an individual site's NPS webpage is inadequate for various reasons (including facts that it is not the actual source used by editors and that it only selectively reports some NRIS info.)

It is notable that regular DYK editors, whom I believe are well-meaning but somewhat misguided when they do this, regularly remove references to NRIS because this reference does in fact suck, in new NRHP articles up for DYK. I could provide several links of this happening. It is tedious and/or embarrassing to try to "train" anyone to keep the NRISref reference in, despite it obviously sucking.

Anyhow, I would like to try again to fix this up better. --Doncram (talk) 00:15, 22 September 2019 (UTC)


 * What should an improved reference show? I guess i think it should show something like:
 * National Register Information System database, downloaded version of November 2, 2013, provided by the National Park Service.
 * or with another of very few valid version dates
 * or possibly with an optional accessdate field, adding display of something like "accessed September 22, 2019", as the date when an editor got the data from the NRHP infobox generator tool (which currently reports data being in version 2013a, i.e. the version of November 2, 2013, although technically it should be some version 2014a instead and display September 30, 2014, instead, i think).
 * Any information about how someone could download their own copy of the NRIS database (not recommended) should be at the National Register Information System article.
 * And the "refnum=" and "name=" fields should be dropped; any current usages should be put into an error category to be fixed. Because it is simply not possible to link to the specific data available in the NRIS database.  Those usages are linking to something else, the NPSFocus system;  the usages could be "corrected" to link more explicitly there (although frankly "there" is a horrible place to link to, so IMHO just dropping those fields would be better, but anyhow for discussion here let's just assume they can be linked to there, which is just something different than NRIS).
 * Okay? --Doncram (talk) 07:33, 22 September 2019 (UTC)

NRHPweekly
NRHPweekly is a related reference, in draft form now, which is intended to facilitate referencing to the National Park Service's "Weekly" announcements of NRHP listings. Information which will eventually get included into an NRIS database version, but is not itself the NRIS database. See its Talk page about some coding stuff, which was originally intended to be an expansion of this NRISref template. I think it is different, best kept separate, but coders should be aware of it. --Doncram (talk) 02:59, 22 September 2019 (UTC)

Certificate errors
When you click through the link created by this template, the resulting web site is giving certificate errors. See the first reference at Norwich City Hall (Connecticut) for an example. I sent a note through the nps.gov web site asking them to fix it, but someone more familiar with the content of the NRHP web sites might want to look and see if there is a better URL that should be used by this template. – Jonesey95 (talk) 03:46, 7 January 2020 (UTC)
 * , see Wikipedia talk:WikiProject National Register of Historic Places/Archive 68. Looks like the discussion died as unresolved. Can you help? MB 04:18, 8 January 2020 (UTC)
 * It appears that the old site no longer exists, and that someone needs to figure out where the links should go. Even if I accept the security risk and click through, I am not taken to a site that makes it clear to me how to find the information I seek. Where should readers end up if they click the links that this template generates? Paging and  and  as well. I am happy to edit the template, but I don't have enough NRHP expertise to figure out what to change it to. – Jonesey95 (talk) 05:13, 8 January 2020 (UTC)
 * If I understand matters correctly, specifying the refnum parameter will invoke, which takes you to the correct location. The problem is the default case without refnum. I don't follow the use case for using this template without the reference number, and I don't see how a user would know which date to use. Anyway, the closest thing to a landing page with the new system might be https://npgallery.nps.gov/nrhp. Mackensen (talk) 12:39, 8 January 2020 (UTC)
 * It looks like without the refnum, we could default to https://npgallery.nps.gov/nrhp/Download?path=/natreg/docs/All_Data.html or to https://npgallery.nps.gov/NRHP. Any preferences? I'll set it to the latter for now, since it's a search page, which seems to be the intent of the link. – Jonesey95 (talk) 14:57, 8 January 2020 (UTC)


 * It is a long-standing horrible problem that the very common template directs readers towards themselves downloading the NRIS database, which they absolutely should not do. I suppose the problem is failure of NRHP regular editors plus occasional hangers-on to be able to agree on anything.  The NRISref template has been a magnet for, well, idiotic ideas.  Keeping the certificate error would perhaps be better! --Doncram (talk) 15:26, 8 January 2020 (UTC)


 * I agree with Mackensen that it would be more useful to use the refnum by default. I've never understood when we haven't always done that. Even though it may not lead to the nom form, in many cases it does. This is especially useful in cases where the article doesn't have another ref to the document (all the NRIS-only articles). I have learned to right-click on the number in the infobox, but the average reader may attempt to use the ref in the reference list and get nowhere. With the change Jonesey made, you are taken to a search page. If you enter the refnum there for a property where the form isn't available there, you get the "form has not been digitized" image. But if you right-click on the refnum in the infobox you get the same "form has not been digitized" image plus a dump of the metadata that shows where the info in the infobox came from. I think that is better than nothing. I would support changing the template so it creates a reference in the same way. MB 15:53, 8 January 2020 (UTC)

I've modified the template to add a tracking category for pages which invoke the template without a reference number: Category:Articles using NRISref without a reference number. It will take some time to populate. Mackensen (talk) 17:06, 8 January 2020 (UTC)
 * , as far as I can tell, if the template does not have a refnum, it can't take the reader from a WP page to a specific NPS page with information related to the historic site in question. Currently, if I don't know the refnum and click the first ref at the bottom of Norwich City Hall (Connecticut), I can then type "Norwich" and get a short list of results. I don't have to download a whole database. I think this is an improvement over a scary security error that most readers won't understand, but I have been wrong. I also think that including a refnum where possible would be a good improvement. – Jonesey95 (talk) 20:34, 8 January 2020 (UTC)
 * , Your change is a definite improvement over the scary security error. The bigger issue I am talking about, and I think Mac too, is that in the normal use of this template, inside the NRHP infobox, the template is not normally invoked with the refnum. Rather than editing all 70k articles, is here any technical way to accomplish this automatically? When this template is used within, can it "see" the refnum which is a parameter to that template? MB 00:00, 9 January 2020 (UTC)
 * I don't think this template needs to be used within Infobox NRHP. That infobox links directly to the NPS Gallery page using NRHP Focus, so an additional reference in the refnum parameter, as in Norwich City Hall (Connecticut) is redundant (and, without a refnum parameter in the NRISref template, not useful). If the new tracking category shows many placements of this template within the infobox's refnum parameter, we can probably recruit a bot to strip them out. (Edited to add: Out of 70,835 transclusions of this template, it appears that 70,648 of them do not use a reference number.)
 * For those who don't mind looking at a little code, here is the beginning of the refnum section of Infobox NRHP:
 * For those who don't mind looking at a little code, here is the beginning of the refnum section of Infobox NRHP:


 * label22 = NRHP reference #
 * data22 = {{#if:{{{refnum|}}}|{{#invoke:String|replace|{{{refnum}}}|^%s*(%d+)|[{{NRHP Focus|%1|url=yes}} %1] ...
 * In short, it says "if refnum= is populated, grab the reference number and use it to make a link to the NPS Gallery page for that refnum." When making the link, it will ignore any text after an initial string of numbers, so if there is also a {{tag|ref}} in the refnum parameter, it will not become part of the link. There is some other processing that happens there as well, but I didn't take the time to parse it. – Jonesey95 (talk) 01:36, 9 January 2020 (UTC)
 * The refnum link in the infobox is an external link but it doesn't put anything in the reference section. It may be functionally equivalent, but in practice people expect the article's references to be down in the reference section. So I don't think it is redundant. If we were to have a bot remove them all, we would be left with over 1000 articles with no references. Then they would eventually get tagged with "This article is completely unreferenced". Not to mention that the ref is named "nris" and often used elsewhere in the article. Those would all become cite errors. Norwich City Hall (Connecticut) does that. MB 01:58, 9 January 2020 (UTC)

, Can you answer if it is possible to somehow get the refnum from the superordinate when this templated is nested within it? MB 00:24, 31 January 2020 (UTC)

Article transclusion
When editing Erie Canal and previewing without any changes, the article (Erie Canal) is shown at the top of "Templates used in this preview". That is due to use of NRISref. If I edit a non-existent article such as REDRED and copy/paste  into the edit box, then preview, it shows it is transcluding REDRED. I had a quick look but I can't see where that is happening. For my curiosity, what is going on? Johnuniq (talk) 04:13, 18 January 2020 (UTC)
 * That has nothing to do with this template, specifically. The various citation templates in the Citation Style 1 family call the article to check to see if use mdy dates or use dmy dates are included in the article. If so, the citation template renders the publication, access and archive dates appropriately. Now this template does use cite web, which is one of the CS1 family, so an article using this template will exhibit this behavior.  Imzadi 1979  →   05:03, 18 January 2020 (UTC)
 * So NRISref called a citation template and it did the transclusion! OMG. Thanks. Johnuniq (talk) 06:52, 18 January 2020 (UTC)

fix template to drop linking to any URL, and to properly convey the source is an offline database

 * See also Template talk:NRHP row

The NRISref template has been modified by editors who were too hopeful/optimistic that it could link to online, refnum-specific information. There has been massive, repeated misunderstanding over more than a decade about what the NRIS database is.

(Including within Wikipedia_talk:WikiProject_National_Register_of_Historic_Places/Archive_48, the October-November 2010 discussion that led to creating this NRISref template and led to a huge bot run, fixing 22,000 or 36,000 or so references to recently dead interface http://www.nr.nps.gov/. What was implemented then correctly included no URL link, as far as I recall;  some discussants then advocated adding, and others arriving since have repeatedly added, refnum-specific or non-refnum-specific URLs which NEVER EVER have worked as intended. --Doncram (talk) 22:35, 17 December 2020 (UTC))

It is in fact a downloadable big database, which has, in various versions, been downloaded by various NRHP editors including myself. It is IMPOSSIBLE to link directly to the database's complete set of information specific to one refnum. It is also impossible to link, in one URL, to the partial information available for each of the refnums relevant to an NRHP-listed site having multiple refnums. (revised --Doncram (talk) 17:51, 17 December 2020 (UTC))

In the past for some long periods of time the template has linked to an NPS webpage where the whole database could be downloaded, which I think may no longer be available at the URL given, and which at any rate was extremely poor in serving Wikipedia readers: no one could possibly think it would be a good idea for general readers to download the database themself.

The version in place only sort of works for linking to refnum-specific information, for some cases, by linking to a refnum-specific url within the https://npgallery.nps.gov/NRHP system.

1) For many articles now, it fails to link to a "correct" location that way. Charlestown Historic District is the article I am currently looking at, where it fails. Sorry, Charleston Historic District was what I meant; clicking on the footnote links to https://npgallery.nps.gov/NRHP.  Note that there are about SIX reference numbers associated with the district; linking to just one refnum's NPGALLERY page would be incorrect. --Doncram (talk) 17:35, 17 December 2020 (UTC)

2) For all the articles where it "succeeds", it is linking to an incomplete NPS webpage about the given refnum, with only a few, selected NRIS fields, which is in almost all cases NOT the source used by NRHP editors. In almost all articles, nrhp editors used all or part of the full database, perhaps the latest version of database made available by one NRHP editor at a private webpage, refered to as the "NRHP infobox generator".  (See wp:NRHPhelp for info about that.)

I request that the current sandbox version, https://en.wikipedia.org/w/index.php?title=Template:NRISref/sandbox&oldid=994799698, which uses "cite report" rather than "cite web", and which provides no URL at all, be put into place here, thereby restoring correct information. --Doncram (talk) 16:44, 17 December 2020 (UTC)
 * I see that I have been pinged in a related discussion. I'm willing to help, but is the link really not useful? If I go to Mabel McDowell Adult Education Center, scroll to the NRIS reference, click through to the search page, and search for "Mabel McDowell", I see a list of entries that includes the entry for this school and a link to download the PDF. That seems useful. It would be even more useful if we could link directly to it, but I don't know if the NPS provides reliable identifiers that are permanent. – Jonesey95 (talk) 17:07, 17 December 2020 (UTC)
 * We can and do link to the NRHP and/or NHL nomination documents for sites. The two PDF files, one being the text National Historic Landmark nomination document written by author Laura Thayer and one being the accompanying photos, are in fact explicitly used as sources in the Mabel McDowell Adult Education Center article.  The PDF files are not part of NRIS.  It is NOT helpful to conflate the NRIS database with the NPS Gallery page which happens to provide links (in this case) to the PDF files and also to echo part of the NRIS database info for the site.
 * I'm sorry, no offense intended, but failure to understand that the NRHP and NHL documents are linkable, and in fact are already linked in most cases, is kind of an indicator that the NRISref changes made in 2019-2020 were not well enough informed.
 * I see in discussion section above that Jonesey95 was encouraged by other(s), and the failure to understand what NRIS is (an offline big database having 50 or 100 fields or whatever) is quite general to many previous editors arriving here. --Doncram (talk) 18:52, 17 December 2020 (UTC)
 * I freely admit that I do not understand the nuances here. I am a template editor, and generally edit only things that I understand, or things that appear to have consensus from a group of editors who appear to understand what is wanted. If I have erred in my understanding or in implementing consensus incorrectly, I apologize.
 * I am stuck, however, on Charleston Historic District, where the statements being supported by the NRISref template are all visible on the NPS pages linked from the refnum links. Is it the use of the NRISref template that you are objecting to in this case, since the statements are being supported by documents and web-based information available at the NPS website rather than the NRIS database itself? If so, I think I get that, and maybe a different template should be used in that case. Perhaps all NRISref templates with refnum could display text other than "NRIS" as the source, or they could be changed to use a different template. Just throwing out ideas here.
 * Removing a link to a page that directly supports the information cited in the article seems contrary to WP:V, which is where I am getting stuck. I think this is a productive conversation, and I am open to any option that preserves WP:V and has consensus of the NRHP project. So far, it's just the two of us talking here, which is not ideal for consensus-building and drawing on the knowledge of HP-interested editors. – Jonesey95 (talk) 19:55, 17 December 2020 (UTC)
 * Hmm. Pinging User:MB too, and I did give notice about this at wt:NRHP.  Adapting the perhaps-clever refnum-specific coding for usage in a different template, say a new template:NPSgalleryref (currently a redlink) could be a good idea (I think that such a template might exist or have previously existed, although I am not finding it in Category:National Register of Historic Places templates).  But the concern here is WP:SAYWHEREYOUGOTIT, the requirement within Wikipedia content guideline Citing sources that editors must cite where they got their info.
 * A new "NPSgalleryref" template with refnum-specific coding could be created and recommended for use by NRHP-type editors with big warning statements that it should be used if and only if that is the actual source used by them. And it could be clarified in documentation there that NPS Gallery, as of 2020 and for some preceding years, is an interface which serves up links to NRHP and NHL nomination documents and which provides partial NRIS info for a given refnum (e.g. the area descriptive info field "An area roughly bounded by Broad, Bay, S. Battery and Ashley and an area along Church bounded by Cumberland and Chalmers" within what the NPSgallery webpage for 66000964 provides), and that it probably provides some info which is not part of dNRIS too, such as "Asset ID:4b343bb8-a9df-4ede-a6a7-60298532368b" and "Related Portals:NPGallery Main Search".  It must be noted that NPS Gallery is an interface which maybe is temporary, may go away entirely.  It would be okay for references to exist linking to it, even if the link goes dead, that happens all the time.  And then the NPSgalleryref template itself could be centrally edited to clarify that while the specific link cited originally no longer works, much of the info it presented is ever-verifiable as part of the fundamental NRIS database.
 * But it was and would be wrong to mass-convert all NRISref usages which have been edited to include a specific refnum in their call, to retrospectively assert that NPS Gallery rather than NRIS was the source. In fact, as can be confirmed by prolific NRHP article creators such as myself and others who might be reached at wt:NRHP, the "NRIS infobox generator" version of NRIS was actually the source used in almost all NRHP infoboxes.  Note, as WP:SAYWHEREYOUGOTIT states, it is okay/good for an editor to cite the actual fundamental source (NRIS) rather than the specific interface they used to get to that source's info;  this is what NRISref should (again) do.  It is okay, but a different thing, for some future users to cite the specific interface NPSgallery, if that is the interface they are using. --Doncram (talk) 21:23, 17 December 2020 (UTC)
 * A new "NPSgalleryref" template with refnum-specific coding could be created and recommended for use by NRHP-type editors with big warning statements that it should be used if and only if that is the actual source used by them. And it could be clarified in documentation there that NPS Gallery, as of 2020 and for some preceding years, is an interface which serves up links to NRHP and NHL nomination documents and which provides partial NRIS info for a given refnum (e.g. the area descriptive info field "An area roughly bounded by Broad, Bay, S. Battery and Ashley and an area along Church bounded by Cumberland and Chalmers" within what the NPSgallery webpage for 66000964 provides), and that it probably provides some info which is not part of dNRIS too, such as "Asset ID:4b343bb8-a9df-4ede-a6a7-60298532368b" and "Related Portals:NPGallery Main Search".  It must be noted that NPS Gallery is an interface which maybe is temporary, may go away entirely.  It would be okay for references to exist linking to it, even if the link goes dead, that happens all the time.  And then the NPSgalleryref template itself could be centrally edited to clarify that while the specific link cited originally no longer works, much of the info it presented is ever-verifiable as part of the fundamental NRIS database.
 * But it was and would be wrong to mass-convert all NRISref usages which have been edited to include a specific refnum in their call, to retrospectively assert that NPS Gallery rather than NRIS was the source. In fact, as can be confirmed by prolific NRHP article creators such as myself and others who might be reached at wt:NRHP, the "NRIS infobox generator" version of NRIS was actually the source used in almost all NRHP infoboxes.  Note, as WP:SAYWHEREYOUGOTIT states, it is okay/good for an editor to cite the actual fundamental source (NRIS) rather than the specific interface they used to get to that source's info;  this is what NRISref should (again) do.  It is okay, but a different thing, for some future users to cite the specific interface NPSgallery, if that is the interface they are using. --Doncram (talk) 21:23, 17 December 2020 (UTC)
 * But it was and would be wrong to mass-convert all NRISref usages which have been edited to include a specific refnum in their call, to retrospectively assert that NPS Gallery rather than NRIS was the source. In fact, as can be confirmed by prolific NRHP article creators such as myself and others who might be reached at wt:NRHP, the "NRIS infobox generator" version of NRIS was actually the source used in almost all NRHP infoboxes.  Note, as WP:SAYWHEREYOUGOTIT states, it is okay/good for an editor to cite the actual fundamental source (NRIS) rather than the specific interface they used to get to that source's info;  this is what NRISref should (again) do.  It is okay, but a different thing, for some future users to cite the specific interface NPSgallery, if that is the interface they are using. --Doncram (talk) 21:23, 17 December 2020 (UTC)
 * But it was and would be wrong to mass-convert all NRISref usages which have been edited to include a specific refnum in their call, to retrospectively assert that NPS Gallery rather than NRIS was the source. In fact, as can be confirmed by prolific NRHP article creators such as myself and others who might be reached at wt:NRHP, the "NRIS infobox generator" version of NRIS was actually the source used in almost all NRHP infoboxes.  Note, as WP:SAYWHEREYOUGOTIT states, it is okay/good for an editor to cite the actual fundamental source (NRIS) rather than the specific interface they used to get to that source's info;  this is what NRISref should (again) do.  It is okay, but a different thing, for some future users to cite the specific interface NPSgallery, if that is the interface they are using. --Doncram (talk) 21:23, 17 December 2020 (UTC)


 * The example in the template documentation also links directly to a page with a link to a downloadable PDF. Are you requesting that this link be made to stop working? I am sure that I misunderstand this request. – Jonesey95 (talk) 17:10, 17 December 2020 (UTC)
 * Sorry, i corrected above my link, was speaking of "Charleston Historic District" not "Charlestown Historic District". Providing access to a downloadable PDF does not sound like a bad thing to do, but that is not the point of NRISref, which is to indicate the source of the database-type info used in an article.  The PDFs are not part of the NRIS database; the NPSgallery pages happen to include (in many cases but far from all) NRHP nomination PDFs as well as part but not all of the NRIS fields applicable for a given refnum.  Any article should indeed provide explicit link to the relevant PDFs, as most NRHP articles do, including the Charleston one, as explicit sources used in the article.  NRISref is a reference to the offline NRIS database, instead, which does not include the PDFs.  It includes 40 or so fields (if I recall correctly), many more than are reflected in any NPSgallery webpage.  To be clear, the NPSgallery webpage for one reference number, 66000964, of the Charleston Historic District includes:

<PRE>Title:Charleston Historic District National Register Information System ID:66000964 Applicable Criteria:ARCHITECTURE/ENGINEERING Architectural Styles:GREEK REVIVAL OTHER FEDERAL Architects:Mills,Robert Areas Of Significance:ARCHITECTURE Periods Of Significance:1700-1749 1850-1874 1800-1824 1750-1799 1825-1849 Resource Type:DISTRICT Related Collections:National Register of Historic Places Collection Resource Format:pdf File Size (bytes):589943 Date Published:10/15/1966 Parks: National Register of Historic Places Locations: State: South Carolina County: Charleston County Charleston ; An area roughly bounded by Broad, Bay, S. Battery and Ashley and an area along Church bounded by Cumberland and Chalmers Rating: Categories:Historic Asset ID:4b343bb8-a9df-4ede-a6a7-60298532368b Related Portals:NPGallery Main Search </PRE>
 * That is just a subset of info in NRIS for this site. NRISref is meant to refer to the whole database, which includes, for example, size of district (as of 1978 version) being 502 acre.  The NPSgallery page does not provide that field for any reference number.  The reference is meant to allow editors to "say where you got it from".  Changing 60,000 references, from actual references correctly referring to the whole database, to instead refer to something else, in effect introduced about 60,000 errors into Wikipedia. --Doncram (talk) 17:44, 17 December 2020 (UTC)
 * I don't see the size of the district referred to in the article or in any references. Do you have examples from articles where an NRISref template is being used to support claims that are not shown on the corresponding NPS page? – Jonesey95 (talk) 19:55, 17 December 2020 (UTC)


 * Sure. Start at the very beginning of List of RHPs, within National Register of Historic Places listings in Autauga County, Alabama listings:  Bell House (Prattville, Alabama).  The 1 acre size of the property, given in the NRHP infobox, is supported by NRIS reference.  (It has been deliberate practice to state the NRISref reference once in the infobox, attached to one of the last fields, that of refnum, as supporting all the information in infobox unless otherwise specifically supported by a different reference.)  And the second item, Daniel Pratt Historic District, where 140 acre size of district is similarly supported.  The NRIS database is in fact the source "where you got it from" for almost all NRHP articles' infoboxes, as proven in most cases where the NRHP infobox includes information only available from there, not available from an NPS Gallery page.  And in fact the actual source for NRIS info in other NRHP infoboxes, too.  It is rare, vanishingly rare, that an NPS Gallery page is the actual source used by an editor.  If the NPS Gallery page is in fact ever used, then a different reference specific to that could/should be used.  This template is for usages of NRIS database, however. --Doncram (talk) 20:16, 17 December 2020 (UTC)
 * The size field is provided in most NRHP infoboxes, certainly tens of thousands of them. In the article I accidentally linked to, Charlestown Historic District, the infobox reports 49.2 acres, sourced in fact from NRIS, that info not being available at the NPS Gallery page for it. --Doncram (talk) 22:18, 17 December 2020 (UTC)
 * Very helpful, thanks. Can you think of a way to encourage other interested and affected editors to get involved in this discussion? – Jonesey95 (talk) 21:15, 17 December 2020 (UTC)


 * Sorry to keep replying to my own replies, but I added a refnum to the NRISref template at Charlestown Historic District, and it links directly to a useful page about the historic district. Is the problem that we need someone to fix the articles in ? – Jonesey95 (talk) 17:14, 17 December 2020 (UTC)
 * Simple answer: No. --Doncram (talk) 17:46, 17 December 2020 (UTC)
 * And I would delete the "|refnum=" field and any content, and delete that error-type category; the field and a corresponding error=type category however could be used in a new template:NPSGalleryref, however. --Doncram (talk) 01:00, 22 December 2020 (UTC)
 * I wonder if it would be acceptable to remove the link by default, and then, if the refnum is present, to add something like "Some information may be available at refnum". – Jonesey95 (talk) 01:05, 18 December 2020 (UTC)
 * , I haven't gone through all this text today, but there was a related question above that you probably didn't see almost a year ago because of a failed ping. Every NRHP infobox already has a refnum in . Since this template is nested within the infobox template, is there a way to get the number automatically? If so, that would be a step in the right direction. <b style="color:#00FF00">MB</b> 02:22, 18 December 2020 (UTC)
 * The NRISref template can't see the infobox's refnum parameter unless it is put into the infobox template itself (not into a transclusion, but into the code in the Template:Infobox NRHP page). That is not normally how things are done; it's usually better to build the functionality directly into the infobox code. – Jonesey95 (talk) 14:15, 18 December 2020 (UTC)
 * Well, I think it would not be appropriate to build a reference to NRIS into all NRHP infoboxes. Because many NRHP infoboxes are created by editors copy-pasting and putting in what they can, drawing on different sources than NRIS.  The "NRHP infobox generator" access to NRIS which has been convenient for many NRHP editors to use, in fact only works for NRHP places listed before some date in 2014.  That "generator" does a fine job of providing an NRISref reference "showing where you got it".  I do believe that the National Park Service has an evergreen version of NRIS database which could be obtained at any time by request;  it just happens we have not asked for it and collected it (and if we got a new copy, for that to be very useful then the editor who has kindly provided the "generator" service would have to be asked to update their service).  But NRHP editors create new articles all the time for newer listings.  Often using the National Park Service's weekly announcements of new listings as the source for NRHP listing name and refnum and NRHP listing date.  Or maybe using a state history department's announcement, or a newspaper story.  Overall, there has been good practice in editors using the NRISref reference when NRIS was their source, and in NOT using it when it is NOT the source.
 * Let's say this request is/was 1) about restoring proper presentation of template:NRISref to document usage of the offline database NRIS when that is the actual source used, without misdirecting readers to go anywhere. And to cast off hijacking of this template to (mis)represent anything different.
 * It is okay by me to further call for 2) creating new template:NPSGalleryref, adapting code from current version of NRISref, for specialized usage when the NPS Gallery page for a given site is truly the source used by an editor to get some fields of NRIS data or some other data. This could in fact be valid, could in fact be what some frequent NRHP editors might need, if they are actively creating NRHP articles for listings from late 2014 until recently.  Even though NPS Gallery is likely to go away entirely, it is currently a valid source and would be "where you got it" if in fact it is "where you got it".  Other sources for some NRIS-type data are the full NRHP nomination documents (which might be available from NPS or from NARA (see wp:NRHPHELP), but I think actually some NRIS fields are populated by action of NPS staffpersons who are reading the full NRHP nomination document (or alternative state or local form) and extracting objective info and/or making subjective characterizations based on the documents.  So all of what NRIS will say about a given site is not fully determined in any automatic way from source documents;  NRIS is itself an independent(? is that the right word ?) source and will remain so. Usually the NRHP nomination documentation will not include NRHP listing date or refnum; NRIS is especially useful for those two fields.  (By the way, the NPS Gallery interface includes the refnum, labelled as "National Register Information System ID" and it probably always also includes the NRHP listing date, but labels it as "Date Published", which is so unclear that I consider it unacceptable as a source.  It can only be interpreted as meaning the NRHP listing date if you have a different source providing the NRHP listing date.)
 * (Ping to Magicpiano, who recently commented helpfully in linked other template discussion: could you possibly please comment here too?)
 * --Doncram (talk) 00:17, 22 December 2020 (UTC)

Doncram, there is a lot of text here and I am sure you have some very good points. But I still think until there is something better, we should improve the NRIS ref as long as it is the only source used in many articles. I ran across a new article today, Moynihan Train Hall, that borrowed a NRIS ref from another article. I added refnum to and now there is a working ref in this article that documents a specific fact (a building was listed) and if anyone clicks on the link, they get the image and can download the entire document. This is better than without the refnum specified. In South Main Street Historic District (Memphis, Tennessee), I added the refnum in 3 of the 4 links in the infobox - that was the easiest way for me to find the nomination form. I think we should have a bot add the refnum in every article as a short-term improvement. I realize that in many cases, you will still get "document not digitized" but that is not any worse than what happens now. Is there really any down-side to this. You seem to be saying that the refnum is meant to be the source for all the info in the infobox, but that is an inference that probably isn't obvious to the casual reader - the citation is after the refum so it really only supports the refnum itself. In a developed article, the data in the infobox should also be in the prose and supported there with refs. <b style="color:#00FF00">MB</b> 03:17, 31 December 2020 (UTC)


 * By the way, Atlantic306 just marked "refimprove" on a couple new NRHP articles due to the poor nature of the NRISref reference, which currently (and for a long time) links to a search page. It remains that NRISref should be a non-linked reference giving attribution to the downloaded NRIS database, and that "refnum=" should not be added to instances of its use, because that tricks the reference to imply (falsely) that a specific NPS gallery page was the source for data that came instead from the offline NRIS database in approximately 72,000 cases.  MB's points are okay, but speak to using a different NPSgallery template to be created for use with specific refnums.  It would simply be wrong to claim those were the sources when they were not and do not contain much of the NRIS database data used in articles.
 * I would simply fix this template but am blocked from doing so because of unnecessary and unhelpful admin only restriction of this template. --Doncram (talk) 01:43, 11 May 2022 (UTC)!
 * Can you make the desired changes in the sandbox version and then request review? Mackensen (talk) 01:47, 11 May 2022 (UTC)
 * All right, User:Mackensen, the current sandbox, permalink here no, this revised version of May 12, is what is needed. Here's what it generates when invoked most commonly (when copy-pasted from the current version of the "NRHP infobox generator", which uses database version downloaded from NPS in 2013:


 * In contrast to this current version of the reference (which links to a search page):


 * AFAICT it only could be improved by dropping one period that the code puts in unhelpfully, but it is okay as is. I fixed that. Note this version uses "cite report" with "type=none" to avoid providing any URL. It does link to article National Register Information System which should describe the database. And note it dispenses with nonsense/false statements.
 * However I'm afraid you will not implement this, because you may wish for "consensus" and there is a shortage of informed editors to give support. Unfortunately editors without long experience will likely again comment with wishful thinking in support of achieving impossible links to data that cannot be linked. I would fix this by bold edit myself if i could, and no one would actually dispute the fix because no one really cares or knows enough to obtain a new consensus for a wp:BRD reversion or other change. Some might show up with some complaint or another but would not be able to defend anything other than this version. But no current administrator will step in and make the necessary fix, out of unfamiliarity with the issues and inability to discern which commenters deeply know the facts here. Sorry, i am fatalistic about idea of this ever being fixed. And it is in fact bad that Wikipedia readers (and newer NRHP editors) have long been dis-served by bogus links and misdirection in this reference used 72,000 times. Doncram (talk) 08:02, 11 May 2022 (UTC) 02:04, 12 May 2022 (UTC)
 * User:Mackensen, in the spirit of wp:BRD, could you please just implement the sandbox version. This can do no harm, and it removes 72,000 bad links. --Doncram (talk) 02:11, 12 May 2022 (UTC)
 * My boldness is somewhat tempered by the number of pages :). I've been previewing this change with various pages. Let me make sure I understand the full import of the change. The base link to the NRIS system, which you get if you don't specify refnum, is unhelpful/broken. If you do specify refnum, you get the link to the NPS Gallery, which is a useful subset of NRIS but provides a potentially incomplete reference. This change removes the link altogether (and makes the refnum parameter superfluous).
 * This does degrade an article like Moynihan Train Hall, where I think a well-meaning editor did source information to the NPS Gallery site using this template. I believe I've done the same at times. Before I make this change, what should editors be doing in this case, and do we have a way forward for them? Mackensen (talk) 10:57, 12 May 2022 (UTC)
 * Without looking at the edit which added it, I agree that in the Moynihan Train Hall article that it appears by its presentation that the usage of NRISref is just to support the fact of NRHP listing of the James A. Farley Building. The NRISref reference could be replaced by a new template:NPSgallery as I have suggested, which would also link to the horrible NPS webpage which repeats false statement "The PDF file for this National Register record has not yet been digitized".  However IMHO a different reference altogether should be used, speaking better to which portion of the larger building is National Register listed vs. the new Moynihan stuff.
 * The example of NRISref with refnum= being used suggests possible ignorance about what is National Register listed, because any editor actually using that source is themself not given any useful info. In context, in part because the term James A. Farley Building is not wikilinked when NRHP listing is mentioned, I think the article is suggesting the relatively brand new Moynihan Train Hall is NRHP-listed;  it is not.  The NRHP document which is in truth available here, from 1973, at NARA is all about the spectacular historic exterior colonnades. It is my guess that NRHP listing of the building might properly be mentioned as having complicated the major Moynihan modifications to the building, but everything about the Moynihan portion is partially undermining the NRHP listing of the Farley building (unless perchance they did some good maintenance or support to the colonnades).
 * More generally I would agree to an effort to review all instances (hopefully not more than 7,000 say) where "refnum=" has been used possibly by well-meaning editors, for which a bot run could easily make a work list (but ideally identifying and sorting by editor who made the change and perhaps also the date of change, to help in making judgments). I would participate in such an effort using AWB (by the way I have been considering just embarking on that anyhow, to remove all instances of "refnum=", which I believe I am entitled to do and would help address this mess which has been running on and on and on.  User:MB would see their watchlist light up.). --Doncram (talk) 05:10, 13 May 2022 (UTC)


 * I'm a little late to the discussion, but since I also use this template a lot I might as well chime in. It does look like the NPS Data Downloads page still has a link to download everything in NRIS, if we want to have a link somewhere. That being said, I'm not sure that's the crux of the issue here. I get the sense that most editors using this template, myself admittedly included, are doing one of two things with it, neither of which involves directly accessing NRIS. The first is that they're citing something in NPGallery and are using the template as an easy way to link there; in that case the current version of the template is fine if a little misleading, since it links either directly to the information being cited or to a search page that will get the reader there eventually. The second, more complicated option is they're pulling info from the infobox generator, which pulls data from the NRIS and converts it to an infobox. In that case, saying that the info came directly from the NRIS is still a little misleading, since it was filtered through someone else's code and there's a chance the editor cross-checked it against NPGallery anyway. I don't think linking the infobox generator would serve the reader well either; at least with the NPGallery link the reader can find the information on an official website. The main argument for the NPGallery information being different seems to be that they call the listing date "Date Published", but I think readers can figure that one out even if it's a little jargon-y. TheCatalyst31 Reaction•Creation 01:01, 13 May 2022 (UTC)
 * Thanks User:TheCatalyst31. What you say does not address the fact that the full NRIS database, as of a certain version date, was the source of data in all articles created using the NRHP infobox generator, which I continue to use all the time.  And NRIS includes architectural styles and areas of property that are NOT included the NPGallery info.  It is not true that "at least with the NPGallery link the reader can find the information on an official website".  wp:SAYWHEREYOUGOTIT is clear that which interface one uses to get data does not need to be stated, but the actual source (which is NRIS itself and is clearly not just a webpage in the vast majority of cases) should be cited.  When reviewing instances of "refnum=" having been used by yourself, as you suggest you yourself have often used it as shorthand like a [template:NPSgallery]] link to the webpage could provide, we could take care to substitute that new/different template instead.  This proposal to clarify the NRISref reference to be about NRIS is needed however. --Doncram (talk) 05:24, 13 May 2022 (UTC)
 * P.S. Thanks, TheCatalyst, for providing that link to "Data downloads". I haven't recently been aware of where NRIS could be obtained.  It was at first unclear to me if it offered up the NRIS database, but I see on this page linked from there, about "Geospatial Dataset - (Code: 2210280)" it is stated that the NRIS database is included within the Geospatial dataset.  Mackensen or others may or may not be aware that this NRISref did for a time point readers to an NRIS download page, which was horrible to do to readers. --Doncram (talk) 05:33, 13 May 2022 (UTC)
 * I don't really follow why Doncram says there are 72,000 bad links. There are articles that do not link to any NRHP related reference except this one. See Branch Brook Park and Humboldt Bay Woolen Mill for examples. The NRIS ref gets the reader to a screen where the nom form is available. I would say the the nom form is a vital reference for any NRHP article and ideally every article should references the corresponding nom form. Now I realize that a lot of the forms are not here, but some are and those are of great value in articles that don't link to another copy of the form. I even find articles that do have other links to the form that no longer work, but the NRIS link still works. If anything, I think a bot should add refnum to every transclusion of this template because in some articles, that is the best reference we currently have. <b style="color:#034503">MB</b> 02:49, 13 May 2022 (UTC)
 * Thanks for commenting User:MB. Okay, i don't know how many of 72,000 links link to it, but I believe it is the great majority of them that link to this search page at NPGallery.nps.gov, let's say 65,000 link there.  At one time this might have been helpful to readers seeking information about NRIS, which had a page there which explained a bit about it and where one could download the entire database.  Searching on "NRIS" there now yields error message "Your search returned no results, please try again."
 * The NRISref reference has been partially hijacked, so that now if a "refnum=" field is filled as has been done in at the Branch Brook Park article, then it links to a webpage like this one for Branch Brook Park. Which in this case provides links to the NRHP nomination text document and to the accompanying photos document (which is unhelpful to reader because this duplicates the NRHP document reference in the article, but here appears to be possibly different.  Actually the page provides links to both, twice.  And below it twice provides selected fields from NRIS, but not all of the fields used in the NRHP infobox, for example it does not provide the architectural styles and it does not provide the 359.7 acre of the listing.  So MB, when you edited the article 19 minutes before your posting here, to add "|refnum=81000392", you wiped out the accurate referencing in the article to the NRIS database, and you replaced it with a link to a page which was not the actual source and which does not mention NRIS. wp:SAYWHEREYOUGOTIT is clear that you should reference the source where you got something;  you don't need to say it was obtained from the NRHP infobox generator interface (which it was), but you do need to say NRIS.  I don't mind, MB, I know you mean well, but to be harsh you tampered with the article and you introduced a lie in effect.
 * When User:Djflem created the article in this version in 2012, they did not use that, they used NRIS via the "NRHP infobox generator" interface which yielded more and different information. Another small proof that they used that was that they kept in the term "Late Victorian, Other, French Renaissance" as architectural styles;  the private programmer/provider of the infobox generator has since (properly) changed that to drop the "Other", which was jarring especially when not the last word in such a list.
 * I would most strongly object to a bot run which inserted the "|refnum=" code everywhere. As MB knows, this would introduce horrible (my term) links to NPS webpages like this one for Woodstock (Trenton, Kentucky), which should dismay a reader by its four-times stated assertion that "The PDF file for this National Register record has not been digitized."  That assertion on a NPS webpage is itself a lie, a falsehood, because the article Draft:Woodstock (Trenton, Kentucky) does in fact link to an online digitized version of the National Register document for the place.  The horrible webpage does show selected data from NRIS as before, below.  I and other NRHP editors would HATE HATE HATE that being done, damaging say 65,000 articles created carefully with accurate links to NRHP documents, accurately presented.
 * User:MB, i get your point, that to support just a refnum in an infobox, that a link like that could sort of do. But that is not how the NRIS database has been used, and what you want could be a separate template:NPSGallery or such as I have pointed out previously.  I know you are reasonable and wish that you would agree here for the NRISref reference to be fixed simply as requested by me here, in order to help end the poor state of affairs and to resolve this request outstanding (with small variations) for two years.  If you would agree, I would owe you one and would cooperate fully in some other approach to improving referencing in NRHP articles as you want. Also, BTW, the template:NRHP row needs similar fixing to avoid the bad links to "The PDF file for this National Register record has not been digitized" message when you and I and other NRHP editors have already carefully provided reference linking to the NRHP document PDF files. --Doncram (talk) 04:20, 13 May 2022 (UTC)
 * The "has not been digitized" message only means that NPGallery hasn't digitized the file; it doesn't mean that nobody else has, and I can't see how the NPS could possibly mean it that way when they collaborated with the National Archives to digitize most of the missing files. As for the WP:SAYWHEREYOUGOTIT concerns, that doesn't bind editors to always include the first source someone used to verify data. If NPGallery (or a state website, or the nomination form itself) contains the same information in a more reader-friendly format, there's no reason why we can't cite it instead, and it's certainly not "lying" as long as the information being cited is present in the other source and the second source is as reliable as the first. TheCatalyst31 Reaction•Creation 02:26, 14 May 2022 (UTC)
 * From my experience, NPS itself has digitized lots of NRHP documents for which "has not been digitized" shows at their page, while the document is now available from NARA, and I think I probably could prove that but it doesn't matter does it? The reader doesn't care to learn that one part of NPS asserts that NPS has not scanned something, while in fact it was scanned by Texas or NYS or NARA or by some part of NPS that the one part doesn't know about.
 * I agree that a hypothetical source with a reader-friendly format where the information being cited is present, that it could be cited instead, but NPSgallery pages are not reader-friendly and they don't have all the info from NRIS which is presented in the infobox. But if, say private website NationalRegisterOfHistoricPlaces.com, which is fairly reader-friendly, were also fairly complete and fairly reliable, then I'd be the first to agree to cite that (but we know that it is unreliable, it has errors of several types, and I think we don't like to support them receiving advertising revenue).
 * I hope NRHP editors could agree that we do like readers of the many NRHP articles to be able to follow links to read full NRHP nomination documents; we have invested a lot in developing articles with good references so that they can do so. I hope we could agree that, in an article where we have made such effort, that allowing/encouraging readers to go to a link where they encounter statement that a relevant NRHP document "has not been digitized" doesn't add value and would seem to be negative, a misdirection.  And even if there are NRHP document there, doesn't it seem unhelpful to direct them there, where it is not easy for them to know if that NRHP document is the same as the one cited directly in the article?  In the reference(s) the wikipedia NRHP editor has provided, the citation is proper with authors and dates of preparation that cannot easily be discerned by the reader of the NPSgallery-linked version.  So I think linking there is unhelpful in regards to NRHP documents.
 * Would it help for this discussion to be divided into two parts: what citing should be for NRHP places listed before vs. after 2013, i.e. where the convenient NRHP infobox generator does vs. does not work in getting NRIS data? Or into parts about whether or not the NRHP document is available and cited or not? --Doncram (talk) 15:59, 15 May 2022 (UTC)

URL needs update
Hi, I just noticed that the template url needs to updated from https://npgallery.nps.gov/NRHP to https://npgallery.nps.gov/nrhp. If previous sentence doesn't make a whole lot of sense, please see this page. The link in infobox works fine, but the cited in first line has a url that goes to a 404 page. Thanks. Kiran_891 (TALK) 22:42, 9 April 2022 (UTC)


 * I fixed that particular article by changing the link in the article to point to the specific property (rather than the "landing" page). The default link should be fixed. has updated this before. <b style="color:#034503">MB</b> 23:35, 9 April 2022 (UTC)
 * https://npgallery.nps.gov/NRHP works just fine for me. When I try to go to https://npgallery.nps.gov/nrhp, I am sent to https://npgallery.nps.gov/NRHP. What am I missing? – Jonesey95 (talk) 14:32, 10 April 2022 (UTC)
 * Both links work fine for me using Chrome. But in my Firefox configuration, I get "Resource Not Found - Requested Resource: https://npgallery.nps.gov:443/nrhp/index.htm" with the NRHP (uppercase) link. My Chrome installation is probably default while Firefox may have some additional security features enabled. <b style="color:#034503">MB</b> 14:50, 10 April 2022 (UTC)
 * Strange. The source of https://npgallery.nps.gov/nrhp contains links that point to subpages of /NRHP, so that looks like the canonical path. – Jonesey95 (talk) 15:12, 10 April 2022 (UTC)
 * As has fully been discussed, the reference should not link to any URL. --Doncram (talk) 08:05, 11 May 2022 (UTC)

Template-protected edit request on 27 May 2024
Please remove this text:

|#default = Warning: Template:NRISref used with invalid value for  parameter (help).

The template is out of date and only accepts specific values. The database version should not be required. I'm not sure where at https://npgallery.nps.gov/ I can even find other version dates. Reywas92Talk 18:15, 27 May 2024 (UTC)
 * ✅ I removed the whole switch block because it's only there to check if you entered something that isn't the database version. It seems like the NPS used to publish database downloads every once in a while, but since the 2010s they changed to a searchable database that is continuously updated. SWinxy (talk) 23:24, 27 May 2024 (UTC)