Wikipedia talk:WikiProject Australian historic places/Archive 1

New project
Things to do (in no particular order, feel free to add): Somno (talk) 03:34, 26 October 2008 (UTC)
 * Add parameter to Template:WP Australia
 * Organise category tree under Category:Australian historic places
 * Tag and categorise articles
 * Update to-do list on main WikiProject page
 * Link WikiProject to Australian WP and other relevant projects
 * Add to WP directory
 * Create guidelines to help with new articles
 * Add resources to the list

Clearly
Delineate the differences between 'local', state and federal registers/legislation of what is historic/listed SatuSuro 03:44, 26 October 2008 (UTC)

Project scope
Could I suggest the project scope be expanded to include items on State heritage registers as well as those classified by the Commonwealth and the National Trust?

The Commonwealth register covers the major historic sites. The National Trust listings include a wider range of places but can be a little too focused on historic houses and architecture and miss places of (say) economic historical importance. State Heritage registers have robust selection criteria and their inclusion would not in my opinion open the door for trivialities. There is also overlap between State registers and Commonwealth/National trust ones, so their inclusion would also not result in an unmanageable increase in article numbers.

Any views? Euryalus (talk) 00:28, 28 October 2008 (UTC)


 * I agree. Having had a look at the Victorian Heritage Register, I found that most if not all of the entries were of equal or higher notability to the sorts of articles that the US project handles. Rebecca (talk) 01:01, 28 October 2008 (UTC)


 * Sounds good to me. I'd like the idea of a category structure that relates to this - for example, if a building qualifies for all three registers, they would have three categories. Somno (talk) 03:46, 28 October 2008 (UTC)


 * I'm happy to include State registers as well. As a query, is this limited to buildings? I can think of a few fairly mainstream entries on the SA register that aren't (last remaining grape vines in a region, etc), and some far less mainstream items added to the SA State register as part of an attempt to include "movable" items of cultural significance - the Hills Hoist and Frog cakes, to name but two.


 * Given the name of the Project, I'd suggest it relate only to "fixed" items, including buildings but also any other "place" such as a historic garden, natural feature, historic infrastructure and so on. Hills Hoists in general wouldn't therefore qualify, but a particular Hills Hoist with historic significance (the first Hills Hoist? A famous Hills Hoist?) might do. Frog cakes wouldn't qualify at all, though the Balfours bakery might.


 * These are just my opinions - I think the project sounds like a great idea and don't want to come across as proposing restrictive definitions or being overly concerned about process before I've done any work on the content. As an alternative we could leave the scope a little vage and see what evolves over time. Euryalus (talk) 04:18, 28 October 2008 (UTC)


 * I'm happy with places - the attempt to include movable items was interesting, but not, I suspect, wise. :) In which case, the argument is that the project doesn't cover everything in a heritage register, just a subset, which is fine by me. - Bilby (talk) 04:45, 28 October 2008 (UTC)

Comment on the WikiProject X proposal
Hello there! As you may already know, most WikiProjects here on Wikipedia struggle to stay active after they've been founded. I believe there is a lot of potential for WikiProjects to facilitate collaboration across subject areas, so I have submitted a grant proposal with the Wikimedia Foundation for the "WikiProject X" project. WikiProject X will study what makes WikiProjects succeed in retaining editors and then design a prototype WikiProject system that will recruit contributors to WikiProjects and help them run effectively. Please review the proposal here and leave feedback. If you have any questions, you can ask on the proposal page or leave a message on my talk page. Thank you for your time! (Also, sorry about the posting mistake earlier. If someone already moved my message to the talk page, feel free to remove this posting.) Harej (talk) 22:47, 1 October 2014 (UTC)

WikiProject X is live!


Hello everyone!

You may have received a message from me earlier asking you to comment on my WikiProject X proposal. The good news is that WikiProject X is now live! In our first phase, we are focusing on research. At this time, we are looking for people to share their experiences with WikiProjects: good, bad, or neutral. We are also looking for WikiProjects that may be interested in trying out new tools and layouts that will make participating easier and projects easier to maintain. If you or your WikiProject are interested, check us out! Note that this is an opt-in program; no WikiProject will be required to change anything against its wishes. Please let me know if you have any questions. Thank you!

Note: To receive additional notifications about WikiProject X on this talk page, please add this page to WikiProject X/Newsletter. Otherwise, this will be the last notification sent about WikiProject X.

Harej (talk) 16:56, 14 January 2015 (UTC)

WikiConfererence Australia 2015 - Save the date 3-5 October 2015
Our first Australian conference for Wikipedians/Wikimedians will be held 3-5 October 2015. Organised by Wikimedia Australia, there will be a 2-day conference (Saturday 3 October and Sunday 4 October) with an optional 3rd day (Monday 5 October) for specialist topics (unconference discussions, training sessions, etc). The venue is the State Library of Queensland in Brisbane. So put those dates in your diary! Note: Monday is a public holiday is some states but not others. Read about it here: WikiConference Australia 2015

As part of that page, there are now sections for you to: It would really help our planning if you could let us know about possible attendance and the kind of topics that would make you want to come. If you don’t want to express your views on-wiki, please email me at kerry.raymond@wikimedia.org.au or committee@wikimedia.org.au
 * indicate your interest in possibly attending the conference (this is not a binding commitment, of course)
 * add suggestions for topics to include in the conference: what you would like to hear/discuss (again, there is no commit to you presenting/organising that topic, although it’s great if you are willing to do so), or indicate your enthusiasm for any existing topic on the list by adding a note of support underneath it

We are hoping to have travel subsidies available to assist active Australasian Wikipedians to attend the conference, although we are not currently in a position to provide details, but be assured we are doing everything we can to make it possible for active Australian Wikipedians to come to the conference. Kerry (talk) 00:13, 20 April 2015 (UTC)


 * Folks, just letting you know we will not be proceeding with Wikiconference Australia 2015 originally proposed for 3-5 October 2015. Thanks to those of you who expressed your support. You are free to attend the football finals instead :-) Kerry (talk) 07:42, 3 July 2015 (UTC)

oz historic places tag?
hi, just wondering if there is one? and if not how are articles placed under the auspices of this project? thanks, Coolabahapple (talk) 06:25, 24 April 2017 (UTC)


 * Not that I know of. Most of the sub-projects under WikiProject Australia are pretty much inactive. Last signup here was in 2010. If you want to have a conversation about anything Australian historical, it's probably best to do it at Australian Wikipedians' notice board where you might find a few people interested. There *is* a field in the Template: WP Australia that allows you to tag the article as of interest to WikiProject Australian history which I think is as close as you can get, but I don't know if it gets used all that much. You might want to ask someone like who does a lot of project tagging.

If an Australian historic place is formally heritage-listed, then we do have a category structures for the various heritage registers (and more can be made if needed) and we use within the article but with no special project tag. For non-heritage-listed Australian historic places, I guess you can still use the infobox and there are a bunch of random categories like Ghost towns etc that might be useful.

A lot of heritage-listed articles will get tagged by WikiProject Architecture if they are a fine architectural example of something (most heritage-listings related to a place's role in history or the fact it's architecturally interesting). And depending on whether they are churches, bridges, roads, cemeteries, there are other Wikiprojects who may be interested in those kinds of things. e.g. Talk:St Thomas' Anglican Church, Toowong shows a typical example of this multi-project tagging.

If you have a particular article/topic that's causing you a problem, just say so and I'll attempt to suggest something. But my principle is that I do my best and if someone can come along after me and do better, well that's fine. We are here to work together (even if it doesn't always feel like that).

Kerry (talk) 08:48, 24 April 2017 (UTC)
 * hi, thanks for the info, i will keep it in mind. Coolabahapple (talk) 14:39, 24 April 2017 (UTC)

Convert templates and the NSW SHR generator
Having experimented extensively for the last few days, I'm declaring defeat on automating convert templates in the NSW SHR generator. Althought it sounds easy enough to do, replacing distances and other measurements with convert templates turns out to be very hard to automate. Yes, it's easy to replace "9 miles" with 9 mi but what's not easy is being sure if it is the right thing to do in the broader context and I am now of the view that the broader context is an issue a LOT of the time. Things that complicate the matter are:
 * he travelled 7 to 9 miles each day - needs a range convert template
 * he had a farm 7 by 9 miles square - needs an area convert template
 * the horse travelled at 8 miles per hour - unit is mph
 * 5 miles 6 chains - multiple units used in the measurement
 * He road alongside 5 Mile Creek - proper name
 * Leichhardt wrote "The creek was 9 miles long" (direct quote)
 * the 9-mile post (adjective)
 * 5 miles (8 km) - a manual conversion is present
 * (5 miles) - produces over-parenthesisation
 * 5¾ mile, or 5 3/4 - any use of fractions
 * five miles - spelling out numbers

And with other units, we have other issues, feet and inches can be expressed as 5'6", inches can be abbreviated to "in" which is unfortunately a common English word. I note that when we are talking about gold (as articles about goldrush towns invariably do), we have the issue of the troy weight vs Avoirdupois and frankly it's hard to be sure from the text which weight system is intended, but using a convert template forces us to choose between troy ounce and avoirdupois ounce. It's all very messy and explains to me where there isn't already a gadget/script to do this job.

My conclusion is that you need to see the use of the measurement in context of the larger sentence to make an intelligent decision about whether a convert template is needed, which variant of the template is needed, and what the units actually are. In other words, it's a job for a person. Since the QHR rolled out without any convert templates, I have experimented with using AWB to roll out some use of convert templates. This is actually working well. I put in the "simple rule", AWB shows me what that would do (with sufficient content to see any unintended consequences), I then say Yes Apply The Simple Rule or instead elect to hand-edit the article where I can see the situation is more complicated. I think if we want to automate convert templates for the NSW SHR, then I suggest running some AWB scripts over a large number of rolled-out articles might be the better way (or people can add convert templates manually as they do the roll-out). The risk/return ratio for fully automating it is poor. Kerry (talk) 02:02, 24 May 2018 (UTC)


 * Makes sense. The Drover&#39;s Wife (talk) 07:04, 24 May 2018 (UTC)

NSW SHR photos
I did a preliminary scan of the SHR entries in the NSW HD and found there about 9000 photos (so about 4-5 photos per entry on average). Of those, about 6000 have a copyright holder statement that seems to be for the NSW Heritage Council, NSW Heritage Dept, Office of Environment & Heritage, and variants and misspellings of these, which would seem to indicate they are covered by the CC-BY license. A further group of about 150 photos have no copyright holder statement but indicate one of those same organisation names (or variants) as the Creator, so can reasonably be assumed to also be covered by the CC-BY license. What about the rest? Well, there are many that are clearly attributed to third parties, but there are quite a lot in the "grey zone" where I am frankly unclear what their copyright status is. Some of these have the OEH as a co-holder with another organisation; unclear if these are covered by the CC-BY. Some of these list govt depts that I think are overarching the OEH or the former govt department that maintained the heritage register previously and these are probably included in the CC-BY. There are photos taken by human-named creators for whom the copyright holder is the OEH or similar, suggesting these people are staff members (obviously these are covered by the CC-BY) but there are also photos by the same people, where they are listed as the copyright holder or no copyright holder is listed; these I suspect are owned by their employer and hence CC-BY. Plus things marked as "copyright expired", or that the copyright is "n/a" etc. All in all, it's very messy data for the purposes of deciding if an image does or doesn't fall under the CC-BY license. Open to suggestions here on how to proceed. Kerry (talk) 00:58, 22 August 2018 (UTC)
 * You're generally better at these things than me - don't have much to add except to trust your judgment. The obvious way to proceed would seem to be to do those that are obvious and consider the others at a later date. The Drover&#39;s Wife (talk) 01:07, 22 August 2018 (UTC)
 * The Office of Environment and Heritage (New South Wales) is the OEH. The NSW Heritage Dept no longer exists and its functions were folded into the OEH. The OEH is an agency of the Department of Planning and Environment (New South Wales) (DPE) that was called, at one stage, the Department of Planning (New South Wales) (DOP), and before that the Department of Environment and Climate Change (New South Wales) (DECC). At some age, you may also have come across the Department of Urban Affairs and Planning (New South Wales). The Heritage Council of New South Wales is a statutory body and it's sole function is to make recommendations to the Minister about whether an item should be placed on the NSW SHR. I'm surprised it's claiming copyright ownership of images, as it has no management functions. All of these agencies form part of the Government of New South Wales, who should ultimately hold the copyright status. Rangasyd (talk) 13:34, 26 August 2018 (UTC)

Norma Parker Correctional Centre naming help
I'm struggling to work out what to do with one of these because it's such a significant site, but it doesn't have an obvious title. This site ("Norma Parker Correctional Centre" in the SHR) has variously been the Parramatta Female Factory, the Roman Catholic Orphan School, the Parramatta Girls Home, and the Norma Parker Correctional Centre prison. Each of these (apart from the last one) have articles relating to their specific institutions (all notorious for various reasons), rather than the buildings/site itself. It is on the Australian National Heritage List as Parramatta Female Factory and Institutions Precinct. Anyone have any suggestions?

I also wonder if it might be worth starting with the vastly superior NHL content rather than the SHR on this one. The Drover&#39;s Wife (talk) 08:29, 23 August 2018 (UTC)
 * In such cases, I find it easier to put the HR content into a new Blah Blah building article and link it to the articles about the use of the buildings at the various points in time, which then cross link back to the building article mentioning its HR status. Kerry (talk) 15:07, 25 August 2018 (UTC)
 * I was planning on doing that - but I'm damned if I can work out what to call it. The Drover&#39;s Wife (talk) 21:05, 25 August 2018 (UTC)
 * Clunky though it is, perhaps the NHL "... and Institutions Precinct" might be the best of bad options? -- Euryalus (talk) 21:26, 25 August 2018 (UTC)
 * Hi I hope everyone didn't mind, but I've created Parramatta Female Factory and Institutions Precinct on the basis that naming for all federal heritage register articles should take precedence over the naming for state heritage register articles. If we're happy with what I've done thus far, I suggest that, in this instance, we redirect the above-mentioned articles into the newly created article. We should keep Cumberland Hospital as it is a much bigger site, of which the former female factory is part of. Rangasyd (talk) 17:10, 26 August 2018 (UTC)
 * Having done that and looking further afield I came across this report that gives good context and which may contradict the position taken. (see schematic on page iv) of Rangasyd (talk) 19:06, 26 August 2018 (UTC)
 * I'm not in favour of redirecting the above articles: as Kerry suggested, there are often cases where it makes sense to have one article for the physical buildings and another for the institution (often used with heritage-listed schools) - and this is about the most appropriate case of this approach there is. All of the institutions on the site apart from the modern jail have very notable and very controversial histories and what remains of their buildings today with a brief history of the overall site is a very different subject from their own institutional histories. Rangasyd's additional source is extremely useful, but seems to have done what the SHR did in going with the most recent use of the site for naming (although it's the least notable - though still notable - use of it). I think the Parramatta Female Factory and Institutions Precinct article that Rangasyd has put together is superb and exactly what I was aiming to do for the site - although I'd like to take out the bits that have been taken from at least one of the institution articles (for reasons explained above). The Drover&#39;s Wife (talk) 21:40, 26 August 2018 (UTC)
 * I've reworked the hell out of the article merging in the best bits of all the various CC-BY content available to us. Pretty happy with the result, although it could always use later editing. The Drover&#39;s Wife (talk) 22:58, 26 August 2018 (UTC)
 * Great work, everyone! Having separate articles often works when the place is significant in a number of different ways and therefore likely to attract readers looking for a particular perspective on the site. When I am not sure what to do, I usually go with "make an additional article and cross-link". I figure that leaves the option later to merge if someone can come up with a neat way to do it, whereas it is much harder to unscramble the egg if you have merged prematurely. Kerry (talk) 07:02, 27 August 2018 (UTC)
 * Discussion on this article has been moved to Talk:Parramatta Female Factory and Institutions Precinct. Please refrain from commentating further here. Rangasyd (talk) 15:09, 28 August 2018 (UTC)

NSW SHR priorities going forward
I'm back and think I have finished my data wrangling for Wiki Loves Monuments. So what do people see as the top priority item going forward? Things that I may have vaguely promised to do (in no particular order) are:
 * 1) mini-generator for local heritage content to flesh out the SHR articles as required
 * 2) mini-generator for Register of the National Estate to flesh out the SHR articles as required (and useful for other states)
 * 3) webscrape and upload to Commons the acceptably-licensed photos from the NSW SHR articles, generating commons category wikitext as part of that process
 * 4) generate lists of heritage for each LGA, format to be decided

Or anything else you need or recollect me promising in a moment of weakness. Kerry (talk) 01:21, 10 August 2018 (UTC)


 * Welcome back! How was the trip? I think the top priority would be #1 - the most useful content to fill big voids is the s 160 material from the NSW database (followed by the local content, though it's often more brief). The AHDB is possibly a broader conversation right now, as the NHL and CHL material is both at least as useful as the RNE for topping up SHR content but also warrant future projects in their own right for the ones we don't have at all. The Drover&#39;s Wife (talk) 20:59, 10 August 2018 (UTC)
 * Hope you're both well. My thoughts are:
 * is important, only where it adds value to NSW SHR article.
 * we should not trust content in the RNE. It is dated and meaningless.
 * is a good suggestion. I have been tagging talk pages with photo requested in Sydney/NSW/Riverina/Hunter, etc., as required.
 * is the least important. We should focus our energies and efforts on the NHL, CHL, and each state and heritage registers before tackling local lists. For example, where an item is list on the NSWSHR and is also listed on the NHL or the CHL, then that designation should be added to the article and infobox (if not already added). The priority in the infobox should be:
 * [designation1] UNESCO WHL
 * [designation2] CHL/NHL
 * [designation3] state/territory heritage. Once all state/territory register content is complete, we could move to
 * [designation4] local government heritage. Rangasyd (talk) 12:13, 12 August 2018 (UTC)
 * OK, a mini-generator for local NSW content for use in SHR articles is the winner so I will start with that. Re: the RNE. Don't entirely scoff at it. Someone asked me if I could write Mount Macedon Memorial Cross so I obliged this morning and used (manually) the CC-BY content in the RNE. I used its significance section but didn't suggest that this was the reason for its VHR heritage listing (although it actually says "This place is entered in the Victorian Heritage Register and the statement above has been provided by Heritage Victoria.", so I think I could have asserted that, but obviously that is not normally the case with an RNE entry). As you will see, I use the heading "Significance" in the article rather than the "Heritage listing" heading I use when generating from the NSW SHR. I think this is a reasonable way to deal with content coming from somewhere other than the offical register, use a distinctly different heading name and do not include any commentary that the significance is in any way linked to the heritage listing. Since the CHL/NHL is held in the AHD, the webscraper I write for the CHL/NHL is almost identical to the webscraper as for anything else (e.g. the NRE) that is in the AHD (just as the webscraper for NSW local heritage will be a minor variation of the one used for the NSW SHR). On the photos, I attended a tutorial at Wikimania on using Pattypan which is a tool that takes a directory of photos with a spreadsheet providing the values for the fields (that you normally supply one-at-a-time in the Common's Upload Wizard) for all the photos and then uploads them all for you to Commons extracting information from the spreadsheet. I think (usual famous last words) I should be able to webscrape the photos from the NSW SHR, detect and eliminate those which assert someone other than the Heritage Council or Heritage Dept is the copyright holder (since we can't use those as they aren't covered by the CC-BY license), and then dump them into a directory writing out the row of information for spreadsheet as I go. And then let PattyPans do its magic. Now the photos on the NSW SHR are pretty low-res, but, as always, something's better than nothing. And, again, it needs a human-in-the-loop to actually decide if any of the photos are usable in the article, where to put them, and how to caption them (the heritage register has captions for some photos, but there is no guarantee that it is a useful one). Kerry (talk) 06:51, 14 August 2018 (UTC)
 * And I should also add how impressed I am at the number of articles you guys have rolled out in my absence. Also JarrahTree has also noticed the magnificent effort and made the suggestion that when an LGA was within what most people would consider to be Sydney (which obviously is a slightly rubbery judgement call), that this is added to the first sentence (as in LGA, Sydney, New South Wales) and also in the infobox. This might also be a signal to use the Sydney location map and when we get into the City of Sydney LGA, we might even try using the Sydney central location map. Alas for Queensland, we only have the state map (I guess I will have to learn how to make these maps). Kerry (talk) 07:01, 14 August 2018 (UTC)
 * Yes, I agree on those last two. I've been meaning to add Sydney to the lead sentence where applicable, but I often forget because I've overwhelmingly been doing non-metropolitan articles (at least until now). And Rangasyd and I agreed above about the Sydney location map - but again, I'm a bit prone to forgetting. The Drover&#39;s Wife (talk) 08:59, 14 August 2018 (UTC)
 * 4 being the least important, would need checking for HD syntax, any LGA or Govt Dept gets into the HD with an S170 (I know, you know that, I just learned it) so doing each LGA is one way of focusing, but LGA only HD entries that are not SHR can be left for later. Dave Rave (talk) 07:56, 13 September 2018 (UTC)

Citations I was not expecting to see in the Australian National Heritage List
While working away on my generator for the Australian Heritage Database, I randomly picked Abbotsford Convent as a test case. I was quite surprised to find this citation in it:"Abbotsford Convent. Wikipedia entry. Accessed January 2016. https://en.wikipedia.org/wiki/Abbotsford_Convent" I guess it's a sign of the times! Kerry (talk) 08:57, 15 September 2018 (UTC)
 * So how do we cope with circularity ?  Aoziwe (talk) 11:26, 15 September 2018 (UTC)
 * Well, we don't cite Wikipedia, so generally we have to go to the other article and find the sources in that article and cite them, or eliminate what appears to be supported by that citation. It isn't so much the circularity that interests me, but I was impressed that an Australian Government Department sees Wikipedia as a reliable source. Given that the purpose of a heritage listing is to give legal protection to a place against demolition or re-development, one would assume that there is some importance attached to the quality of the sources that underpin the listing. Or to put it another way, they must think we are doing it right. It's something I imagine we will increasingly see over time as Wikipedia ceases to be the "new kid on the block". Generally you don't see citations to Wikipedia in heritage listings as most new heritage registers create lots of listings at the start but then the rate of listings slows down considerably after that so most heritage register entries were written before Wikipedia existed. The Abbotsford Convent entry is unusual in that regard as it is a recent one (2017). Kerry (talk) 22:23, 15 September 2018 (UTC)
 * Thanks - there is sooo much back story to catch up on.... I suppose if a WP article was a main ref we could hopefully rely on the refs in the WP article rather then the WP article directly (ei, reference inheritance ?)  Aoziwe (talk) 12:02, 17 September 2018 (UTC)

Naming conventions
, what's the deal with titling articles in eastern Sydney at "Thing (Suburb)"? All the QHR entries, all the NSWHR ones that I've done, and nearly all of the random ones in other states are at "Thing, Suburb". I was just a bit confused at you moving Eryldene, Gordon to Eryldene (Gordon) with an edit summary about the MOS, when the convention has always been the opposite. Beyond the convention, it also seems a bit unclear to me: in every Wikipedia naming convention that uses a comma that comes to mind, the bit after the comma refers to a place, whereas brackets disambiguation can refer to anything - Gordon could be an architect, owner, etc. The Drover&#39;s Wife (talk) 13:51, 23 September 2018 (UTC)
 * Fixed all. Thanks :-) Rangasyd (talk) 14:20, 23 September 2018 (UTC)

Confused people at it again about copyright
See Village pump (policy) and Talk:Towrang Convict Stockade. Le sigh. The Drover&#39;s Wife (talk) 08:55, 9 July 2018 (UTC)
 * And I've got someone on my case again. Wentworth Mausoleum, Wentworth Memorial Church, and Salisbury Court (Rose Bay). Rangasyd (talk) 17:06, 19 August 2018 (UTC)
 * , FYI, these are licensed under CC-BY, not CC-BY-SA. The difference is actually critical as we discovered the first time this happened. The Drover&#39;s Wife (talk) 01:11, 21 August 2018 (UTC)
 * You might like this diff that I initiated at the New page patrol instructions, given that what was written there before was an Instruction to remove *all* copyright material without any mention of compatible licence. While it won’t completely solve the problem, at least the NOPs who actual read the instructions might become a bit more alert to it. By the way, the NPPs do see the edit summary before opening the article so mentioning the cc-by in the edit summary will help (I realise you mostly do that just as I do, but I just want to reinforce the value of doing so.) Kerry (talk) 15:43, 5 October 2018 (UTC)

New drafts
- sorry to be a pain, but is there any chance these updated drafts could all go in a new folder? I'm finding it pretty confusing having three near-duplicates of every article, especially because I usually work by scrolling through them in Google Drive from one to the next. The Drover&#39;s Wife (talk) 11:17, 8 October 2018 (UTC)
 * Ironically, that's how I originally uploaded the files and then (very tediously) integrated them into the drafts folder, thinking you needed them all in the one place. I don't think I can unscramble that egg. But reading the documentation suggests that if I set the drafts folder to be read-only by you, you should see only the latest version. So I have done that. Let me know if that has worked. Kerry (talk) 19:41, 8 October 2018 (UTC)
 * alas not :( The Drover&#39;s Wife (talk) 02:12, 9 October 2018 (UTC)
 * There is definitely something wrong with the ==See also== section in the draft articles. They have disappeared. The **See also** section usually contained the portal   And the first two categories appear as Category:Category name and all other category names appear correctly as  (the colon is inserted incorrectly in the first version). Rangasyd (talk) 06:49, 13 October 2018 (UTC)
 * Damn, damn, damn. There's a long story behind this, but I will skip it in favour of working on fixing this. I will then have to reload all the drafts to Google Drive. I will advise when it's done. Kerry (talk) 07:06, 13 October 2018 (UTC)
 * OK, I think I have fixed those problems. The drafts are now uploading into Google Drive as I write this, so probably ready by the time you see this message. Kerry (talk) 15:36, 13 October 2018 (UTC)

National Heritage List
I think we need to have a discussion about using the National Heritage List, and I've been meaning to point this out for a while, but having stumbled across editors at Wikipedia talk:WikiProject Western Australia angry about West Kimberley I think now would be a good time.

I don't think it's a good idea to roll out the National Heritage List universally in the same way as either the state heritage registers or the Commonwealth Heritage List. The National Heritage List has a tendency to group larger areas together in a way that doesn't necessary map well onto Wikipedia or for our purposes. Although this can be useful at times (as at Parramatta Female Factory and Institutions Precinct, where I was having trouble organising content), there are other areas where we should tread carefully. I feel like West Kimberley is one of the latter - it's a big, significant, and sensitive region, and forming an article based on the CC-BY text means we've got a very long discussion of its heritage values plonked on the encyclopedia. The Drover&#39;s Wife (talk) 20:49, 13 October 2018 (UTC)


 * as a west Australian editor, I would not say 'anger', the reaction is with 'some concern' for the validity of material in the Western Australian project - we never have articles that size, through community consensus we usually keep things within a reasonable size, and we check both with the community, and the local pre-existing articles before adding something like that.


 * In the West Kimberley article the bibliography and the current section 9 are well outside of normal heritage issues, and in my opinion should not be considered as anything to do with the usual heritage issues encountered by this particular project.JarrahTree 23:54, 13 October 2018 (UTC)


 * There's no need for a turf war. It comes up here because it's listed on the National Heritage List. It isn't an assertion of ownership over the area. Clearly, there are (serious) issues with the article as it currently stands, and it would be helpful if people got together in one place to work out what to do about it rather than sniping in their own areas. The Drover&#39;s Wife (talk) 00:45, 14 October 2018 (UTC)


 * misunderstanding - it is basically not a heritage article, and most of it is not something that is relevant on this project - it actually is about content in the WA project. rather simple.  no turf, no war - simply a recognition of what consistitutes most of the article.  no sniping, it is simply being very clear about the major issue arising from the article.  JarrahTree 00:58, 14 October 2018 (UTC)

Create list-articles for Queensland historic sites
Several from here participated in recent wp:AWNB, where it was eventually suggested that discussion should move to here.

Perhaps creating a comprehensive list-article, with sub-list-articles as necessary, covering all the 1,752 or so [state-level significance] listings on the Queensland Heritage Register might be the next reasonable step.

There already exist:
 * List of heritage listed buildings in Rockhampton (not table-ized)
 * List of heritage sites in Townsville City, Queensland (not table-ized)
 * List of sites on the Queensland Heritage Register in Warwick (sortable table)

Could this be discussed here? If so I would suggest dividing discussion into five topics: --Doncram (talk) 20:06, 13 October 2018 (UTC)
 * Q-1. [scope and] naming of list-articles (limit to state-level significant ones only, or not, and are these called "buildings" or "listings on the Queensland Heritage Register" or what, etc.)
 * Q-2. format for such list(s) and their tables (to include discussion at least of adopting templates like was eventually implemented mostly by User:Multichill within the U.S. NRHP list-article system, though intending to achieve a different look and feel. Should review some example Featured Lists including in the U.K.)
 * Q-3. partition of Queensland (which could perhaps be by the seven Regions of Queensland, with further breakout of those in Brisbane and various suburbs/cities where necessary to bring down to manageable chunks)
 * Q-4. data/programming/getting it done. (I would hope this could be done by a program generating good presentable articles all at once, better/more efficiently than was originally done for the U.S. NRHP system.  I think that Multichill's comment at wp:AWNB was an offer to do this.)
 * Q-5.future maintenance and tracking (how can updates/additions be handled efficiently, how to mark "accurate as of" status of lists, how to use categories and/or perhaps a linked tracking system like wp:NRHPPROGRESS to support development of articles)


 * Okay, so first things first. I don't think it's a good idea to jump straight to one specific state if the point of this is to do things in a nationally consistent way. There needs to be an overall agreement of Australian editors in all states invested in heritage as to how to do this, otherwise you're going to end up with disputes and a result of a different mishmash down the track. Specifically, WikiProject Western Australia has historically had a lot of opinions about these kinds of things.


 * Firstly, it needs clearly working out whether these are supposed to be lists of buildings on X state heritage register (specific), heritage-listed buildings (includes other heritage lists, but which ones?), or historic buildings (includes other notable buildings as well, but needs to define inclusion). The second and third options necessitate further discussions that need to be had before we go any further. This question is important for whether we can actually get consensus to do a nationally consistent rollout (c.f. the most recent comment in the WP:AWNB discussion).


 * Secondly, most states don't have consistent official regions: Western Australia is an exception to this, as (off the top of my head) is the ACT: what to do about this would be useful to work out now rather than have to switch to something else in some states down the track., what do you think about using the Trade and Investment Queensland regions? The Drover&#39;s Wife (talk) 20:42, 13 October 2018 (UTC)


 * WikiProject Western Australia editors can certainly be invited to participate in a process for Queensland. Doing similarly for Western Australia would certainly be different because partitioning will be different and different formatting can be chosen and so on.  It will easier to do for WA, later, if Queensland is done first.  Queensland looks easiest to do now from data availability standpoint, esp. given yourself and Kerry Raymond and others have created all the individual articles, which creates some further urgency:  those articles should be made accessible to readers through a list-system, shouldn't they?  And it looks easier to partition than WA.  Part of what makes Queensland attractive is that the Queensland Heritage Register search screen shows 1754 places currently with state-level significance, and individual items there show a "date entered" which i am assuming is a registration date.  This is a very well-defined set, and I presume that focusing on these only can be done.  Later on Brisbane's or other local registry's listings could be covered in more detailed list-articles for those areas, perhaps with color-coding on the items listed in the higher Queensland level.  Do you think it would be better to start somewhere else?  I hope it can be okay to agree to start on Queensland, and I hope that it is not necessary to solve everything about all other states, and about Queensland places of lower levels of significance, first. --Doncram (talk) 21:56, 13 October 2018 (UTC)


 * "Starting" before you've worked out what you're starting is an unhelpful approach. So, you put a ton of work into doing Queensland as a big template for everywhere else. Then, other editors in other states then come along and say they reject your whole approach and they're going to do their own thing, and you've got an even worse mishmash than you started with. Discussions to try to find consensus are extremely helpful, especially if you're going to put a ton of work into something. If consensus can be found, then great, you've saved yourself major headaches down the line; if it can't, it can inform a discussion about where to go then. The Drover&#39;s Wife (talk) 23:58, 13 October 2018 (UTC)
 * I know Wikidata can be complex and inscrutable, and expertise in it is scarce, but I really think 80–90% of the effort being proposed here in regards to lists of heritage items could be done there very quickly. The Wiki Loves Monuments lists use header and record templates just like the NRHP ones, but the content is filled from Wikidata using a query which can retrieve any content available and filter it by heritage list, geographic region, coordinate radius, type of object, and many other factors; and can be updated automatically or manually by a bot. Most of the effort spent getting the information into Wikidata by September was linking to existing Wikimedia content such as WP articles or Commons categories to avoid duplication from an import. Now that's essentially done, and 25% of all the heritage lists have been imported and linked, it's a matter of importing the remaining 75%—I intend to do this over the next few weeks/months. The main tasks here are categorising the entries by type of object [instance of] (although they can have just a generic "cultural heritage item"), and working out if very generic names like "House" should use the address as the item name as the English Heritage items do.
 * The advantage of the Wikidata approach is that if lists or templates are updated, the tables/lists can be updated with little effort, so you could add a new field or link to the list in minutes. For example, the Tasmanian list (which is the largest one) is in the form of a PDF(!) on the register's site. Heritage Tasmania also updated the list in September so the list we have is out of date. They have apparently been working on a web service/database for the THR, and let's say they implement it—the query retrieves the register/database ID from Wikidata and adds a formatter URL, and the lists contain this link within minutes. Same with importing updates or linking Wikipedia articles. I would recommend the bot generates the lists onto a project page, and the table template markup be copied to the relevant list article.
 * Now the main issue encountered with the WLM lists, which I believe was also mentioned in the AWNB discussion, is the size and scope of the lists. The WLM lists encountered a template rendering limit in the MediaWiki software of about 300 template calls, so I used LIMIT and OFFSET clauses in the queries to generate several paginated lists. The idea seems to be to break it down geographically, and once the data is in Wikidata it can be analysed to detect any list scopes which could exceed this limit. --Canley (talk) 01:40, 14 October 2018 (UTC)


 * The differences is between state lists is in the understanding of what comprises those lists and where to source them and how the data is arranged, yes WA is vastly different to QLD because WA is segmented and local vs state listing no longer exists they are all considered "heritage listed" their reasons for listing will vary. This was down for a number of reasons, like reducing red tape, consistent approaches across all LGA's, enabling laws to address demolition by neglect, and prosecute for deliberate acts against heritage places. With that went changes to who maintains the registers, how frequently they get updated online. The inHerit website is basically closed with LGA's keeping their own lists, some of are published on line some arent, others were maintained in inHerit for longer than others.  I honestly cant help with formating of other states because I dont their methodology for the data, what ever the methodology copyright doesnt exist on the data only its presentation. Reality is how ever you formulate, divide, or conquer these list they will be different across states and within states, one size wont fit all but they will help guide others. Gnangarra 05:03, 14 October 2018 (UTC)
 * (ec) I just coded the 5 subtopic areas above from Q-1 to Q-5 that I suggest could be used to organize discussion about making list-articles for Queensland's state-level significant historic sites, if we can agree to discuss that. Several comments have been made that could be classified into those subtopic areas and that make progress, e.g. beginning to discuss regions for Queensland (Q-3); suggesting Wikidata could in fact be useful in quickly generating tables (Q-4) and in supporting ongoing maintenance (Q-5).  I'd like to go ahead and break out discussion sections for each of these, and proceed, but it would be better if there was agreement that this could be done.


 * However, the Drover's Wife also suggest above, and in their words at wt:WA, that " it'd be useful to establish if we can find a nationally consistent way of doing this now, rather than ploughing through and having a bunch of arguments down the line". I don't think a nationally consistent way of doing things can be figured out right now, and maybe it never can, and I think it would be too cumbersome to have discussion sections open about all aspects for all states at the same time.  Note you can't discuss regions for dividing list-articles for a state if you don't discuss and use data (e.g. it sure would help to get a "map of all coordinates" generated for Queensland to show where the sites are relative to any partition) and so on.  And you can't discuss formatting if you don't set scope parameters (e.g. are we trying to cover local-level places in a given state, or not), etc. I personally would sorta be willing to try, perhaps setting up sections NSW-1 to NSW-5, VIC-1 to VIC-5, SA-1 to SA-5, TAS-1 to TAs-5, WA-1 to WA-5, NT-1 to NT-5, ACT-1 to ACT-5 and then I would be sorta happy to participate in all 40 separate discussions, if there were others that would too and would have power to make decisions, but I don't think that enough others would participate enough to do anything at all.  I think it would be like trying to force arguments on all matters all at once for all people, including for people who simply are not interested, and it wouldn't work.  I was told at wp:AWNB that the discussion there was too big, and there are other topics that matter for AWNB;  the same applies here, doesn't it?
 * To jump ahead, suppose we had a discussion section TAS-1 about scope and names for Tasmania. What I would say there is that Tasmania appears not to have data available to support identification of the state-level historic resources there.  What I currently understand is that their data is in this big PDF listing of 127 pages, with no indications of different levels of importance.  And there is no search interface and no further information available about any item.  The PDF has maybe 6000(?) items, many of very small importance, based on my small experience trying to find any info at all by Google searching about the 3 fire stations it mentions, and finding nothing, while I believe for many/most state-level importance items in Queensland there is the information already in their 1,752 or so articles, from Queensland, and Google searching would probably provide more info for many/most of them.  I would not support putting any effort into making list-articles out of all the Tasmania items and presenting them as redlinks important enough to have articles, or even as blacklinks worth discussing in the lists.  Can we agree not to discuss how to break Tasmania into regions, and so on?
 * About Western Australia, it is suggested that we should get wp:WA involvement. It's fine to ask there, and for transparency I will post something there referring to this comment, but I don't expect they want to engage in discussion WA-1 through WA-5.  For one thing, data on the state-level significant sites in WA simply might not be available: at wp:AWNB Gnangarra states that WA Heritage is no longer making any distinction between state-level vs. local-level significance.  From working a bit on List of heritage places in Fremantle I believe there is not sufficient data available for the previously-identified-as-local-level Fremantle places to justify listing them in a list-article, and there isn't great info for the previously-state-level ones either.  So any list-article system created couldn't be among Wikipedia's greatest works, so I personally don't want to start there.  Also there is the fact that a Fremantle list exists very saliently, which has problems including that it represents that it covers all levels but really is only a partial list of just the state-level ones, and there was unpleasantness around discussing several aspects of that list already (involving me among others, about scope, about regions, about formatting, and probably more), and I don't think anyone there wants to talk about it or anything much else, unless there was pretty good new reason to do so.  Gnangarra also stated at wt:WA that "The best place to talk about Western Australian topics is at Wikiproject:Western Australia its where you'll find the known to put the whole picture together", which is fine, but doesn't augur well for discussing WA-1 to WA-5 here.  Truthfully, I would be sorta willing to try discussing WA-1 to WA-5 there at wt:WA, but I think forcing discussion there would just be provoking them and cause arguments.  I expect it would be far easier, down the road, for WA editors to engage, IF they eventually saw tangible benefits for Queensland from having a comprehensive list=article system (perhaps measurably greater participation by the public in submitting photos, and readership, and active positive development by multiple editors of the existing Queensland articles from their current form which has been started from the CC-BY text).  Can we agree to put off imposing anything upon WA, please?
 * About NSW, VIC, SA, NT, from what little I know about their data, they seem similar to Queensland, with about 2,000 state-level significance items each. Of course if we worked together just on Queensland we would choose to work in ways that would help for the others.  For example the table templates that I and Multichill have mentioned would be set up to accommodate different color schemes for each, say, and could be reprogramming to allow more major differences of data presentation and order if desired.  But I don't think we need to appreciate all differences up front.  About ACT I have no idea, I don't know how to access any of its info.

I think it just would be simpler for us as human beings with limited cognitive abilities, to just do one state first, and I don't hear any preference for a different state, so could we please try on Queensland, with 5 separate Talk sections somewhat disentangling and guiding discussion, if that is not too much for this Talk page?
 * --Doncram (talk) 06:12, 14 October 2018 (UTC)


 * congratulations on the conflagration of my comments about two different subject areas. It's not provoking to ask questions, its provoking to fail to recognise and allow for differences between systems one key doesnt fit every lock. Gnangarra 09:57, 14 October 2018 (UTC)
 * I am sorry to "conflagrate", in my comment above which is too long and i just collapsed, but we're not disagreeing really, are we? Queensland, data-wise and more, seems doable, and I think Gnangarra is saying go ahead. --Doncram (talk) 15:40, 14 October 2018 (UTC)
 * I think at this point you may as well go ahead and do it for Queensland as a start. It would be a useful thing to have, and it'd be good to just get it done. The Drover&#39;s Wife (talk) 11:04, 16 October 2018 (UTC)

Experimenting with a dashboard
For those who have not seen these yet, dashboards are something which we have started to use with outreach programs to track activity. So I thought I'd experiment with tracking activity of a non-outreach program. Here is the dashboard for the New South Wales State Heritage Register project which is a "category-limited" program (meaning it only tracks edits within a specific category tree, Category:New South Wales State Heritage Register with depth 3 in this case). You have to identify specific editors to be tracked by the dashboard (in this case Rangasyd, The Drover's Wife and me). I note that this dashboard tool is not using any information that is not public from our normal Contribution profiles, but is just filtering it in an interesting way and calculating other metrics from it that we don't normally get to see, e.g. number of words added. The other really interesting statistic which is generated is "article views" which is tracking the number of views on articles in the program (in this case, in the category) *since* the first edit of the article by one of the participant editors to that article. Or to put it another way, the number of views that include any edit done in the program (and yes it will include reverted edits -- the tool is not that smart but for good faith editing it's good enough). This metric is particularly interesting as it is the gift that keeps on giving -- even when a program is complete (no more edits), the article views keep on going up and up and up. The purpose of this metric is to show the newbies in outreach (and their bosses, where applicable) about the multiplier effect of contributing to Wikipedia, or, how a little effort makes a lot of impact. But it occurred to me that regular Wikipedians may not realise the extent of impact themselves, so I thought maybe I should experiment with using it to monitor activity in a non-outreach program. (I note that this dashboard tool is not correctly category-restricting the photo uploads, so it is counting all photo uploads by the participants which is why that metric is much larger than it probably should be for this particular program). Anyhow I am curious what people think about this? Is it interesting? Is it motivating? And I guess whether we should consider tracking more "programs" this way. I note that at the moment, we can only track activity in-total (which I don't think works for "regular" editors like ourselves as we do work across a wide range of areas), by category tree (to a chosen depth, as in this example) or by a template (I have not experimented with this one so not sure I fully understand what that one is doing). Kerry (talk) 22:36, 23 October 2018 (UTC)

Home stretch - Baiame Cave help?
I'm slowly working my way through the needing-a-fair-bit-of-work Singleton entries (which will mean we've done basically everything outside of the three big suburbs in inner Sydney apart from those last few Hawkesbury ones in need of an edit), but I've been putting off doing Baiame Cave because basically the whole article is a series of big quotes and I don't really feel comfortable rewriting a huge amount of text on Aboriginal mythology in my own language. Any ideas on what to do with that one? ? The Drover&#39;s Wife (talk) 06:01, 29 October 2018 (UTC)
 * There is this other entry in the NSW HD about the Baime Cave which is not full of quotes. For the history and description sections, it's no problem to use other content. It's only the heritage listing where we have to stay true to the heritage register. Happy to have a go at it, after I've finished this ibids, which would not be so hard if the heritage register itself wasn't abusing them in the first place. Kerry (talk) 06:19, 29 October 2018 (UTC)
 * Thanks - should've thought of that early. I frequently use other NSWHD instead these days (and am going to need to for several of the still-to-go Singleton ones), dunno why I didn't think it of it on this one. The Drover&#39;s Wife (talk) 06:41, 29 October 2018 (UTC)

New South Wales state heritage items - article generator
This discussion has been transferred from Talk:Rangasyd to this page for ongoing maintenance and input.

I've had to clear a backlog of other things I promised to do for various people, but I started work yesterday on generating articles from the NSW State Heritage Register by web-scraping. I started with the program I used for the Queensland Heritage Register and am progressively adapting it to the vagaries NSW SHR. As with so many things, all states of Australia appear to find it necessary to adopt delightfully different ways to do more-or-less the same thing, sigh! I picked a random heritage site, the Denison Bridge, to use as my initial test bunny. You can see the NSWHR entry for it here and my Wikipedia version of it in my sandbox. As it happens, there is already an article on that topic on Wikipedia called Denison Bridge, so this generated article will never "go live". Now there is a lot still NOT done on this generated article. At the moment, it is wikifying (adding wikilinks) using my Qld wikifier and my architectural elements wikifier. That is why you are not seeing many wikilinks and some of the ones you are seeing are wrong (e.g. George Street links to the one in Brisbane not the one in Sydney). You are seeing some architectural elements correctly wikilinking but even there we do have different architectural styles between the two states and hence there may need to be further development on the wikifier for architectural elements. I haven't yet included the significance statments by criteria but I do have the narrative preamble in place but it needs the last few sentences trimmed (Qld does not have a narrative preamble for significance). The template for NSW SHR citations isn't written yet but will be more-or-less identical with the Qld one -- the NSW SHR uses 2 identifiers (like the pre-2015 QHR did), one for the official registration number (which is just window dressing for our purposes) and one for the database entry number (which is used to construct the URL to the relevant webpage). I also need to construct the CC attribution template but again it will be similar to the Qld one but I will probably parameterise with the 2 numbers (which provides us with some future-proofing against database changes in the NSW SHR).

I am getting to the point of needing "many eyes" to compare what's on the NSW SHR website with what I am generating. Is there important information I am failing to include? Am I picking up rubbish that I could easily exclude/fix (if you read the first sentence of the lede, you see a good example of "garbage in, garbage out" which isn't something I can probably do too much about)? Since when has a bridge been a festival activity?! So I would appreciate any comments you have about this -- might be easier via email (as you can more easily include any screenshots to point out things) - I think you both have my email address (if not, ask).

I am not sure what to do about the citations as the NSW SHR uses parentheses for them (Barker, 1995, 34) but also uses parentheses for other purposes (1900-1920). The QHR mostly used [] for citations which made my life easier. As always with a generated article, a human has to go through and tidy up the things the generator could not figure out. In this case, no machine is likely to figure out what the obvious citation (HO and DUAP, 1996, 88) refers to -- there's nothing in their table of references that matches and indeed, even as a human, I have no idea what this might be refering to.

In this example, there are subheadings present in the History section. Now my generator is actually smart enough to guess these - it warns me when it executes when it sees very short paragraphs as these are often sub-headings. In this case, it said:

Warning: short para Aboriginal people and colonisation. Warning: short para Bathurst: Warning: short para Bridging the Macquarie River: Warning: short para HISTORICAL NOTES ON KEY INDIVIDUALS

but missed what appears to be a subheading about "The Russell brothers and P. N. Russell and Co: " because it is presented as a run-in heading (these are almost impossible to distinguish from normal sentences). My experience is the generator is more likely to be right than wrong about sub-headings, so I might change the generator to emit the suspected subheadings. The human user can always restore them to normal paragraphs if it's the wrong call.

I have not worked out what to do about the images. I did them manually for the QHR using the Upload Wizard on Commons. There may be a better way but I don't know it. Anyhow, that's where I am at. Feedback welcome, indeed wanted. Note the version in the Sandbox may change as I go along so it may not reflect exactly what I write above. Kerry (talk) 22:15, 22 March 2018 (UTC)
 * Firstly, WOW! A huge thank you for putting in the time and effort on how to start this mammoth task. I started doing manual entries but realised that it was beyond the scope of possibility to achieve in a realistic timeframe. Secondly, I hear you about vagaries of each state register. The NSW SHR website repeats a lot of information and some information that is valuable is not included in the places where you would expect it. Thirdly, you chose a great subject to start with. I have walked, run, cycled and driven across the Denison and even swam under it when I was a lad. I knew Theo Barker (cited author) who was, prior to his passing, a local historian. I'm not sure that I have your email address. Although, I'm happy to exchange to facilitate productive discussion. HO = Heritage Office. DUAP = Department of Urban Affairs and Planning (now, Planning NSW). The template for NSW SHR is at cite NSW SHR; unless you're thinking of creating an infobox NSW SHR. Oh, a personal preference, I do prefer use of the relief map as much as possible. Rangasyd (talk) 08:17, 23 March 2018 (UTC)
 * oh wow, this is fantastic! Looking at the SHR article and your sandbox, it'd be good to have the condition information somewhere (though I figure probably hard to do with a bot if there's no infobox field). We might have to take the citations on a case-by-case basis: include the footnotes if we can work out what they mean (I was about to have a crack at tracking that one down and Rangasyd beat me to the punch) but if not just cite the SHR directly as is the case in many of the QHR ones IIRC. It seems like the SHR ones might need a bit more manual editing than the QHR did but that's manageable. Everything else looks good (apart from the obvious things you've highlighted). I very much look forward to seeing this get rolled out! The Drover&#39;s Wife (talk) 09:06, 23 March 2018 (UTC)
 * Thanks for pointing me at the existing template, Rangasyd. So long as the infobox supports the relief map option (I have not checked), it is no problem to generate it. Yes, it is unrealistic to write all the NSW HR articles manually. I also started doing it manually for the QHR and realised I probably wouldn't get the job finished in a lifetime unless I found a better way to do it. While I should be able to get the generated articles a bit better than what you see so far (I have already implemented a couple of the items I previously mentioned e.g. detection of probable sub-headings, and I am actively working on the NSW Wikifier at the moment. TDW, there are lots of fields in the infobox that I didn't use in the QHR articles plus a few you can redefine to suit yourself. So things like the condition may be possible to include there (so long as they are short). Otherwise the condition can go into the body of the article in some suitable place (perhaps with a subheading). I guess my immediate question are what other information to include form the NSW HR web site. Just looking at Denison Bridge, I see the alternate addresses, owner, physical condition, modifications, themes, other heritage listings, which might be candidates for inclusion. I don't think the procedure/exemptions are worth including. Plus, there is the mysterious "further information" which might be worth including but not knowing exactly what one might find in such a field makes it hard to know where to put it in the final article. My immediate goal is to get Denison Bridge to the best we think we can do it to, and then I'll start generating other articles from the NSW HR (which may have different kinds/volumes of information which may make us re-think some decisions we already took). As a general principle, if I can put information from the NSW HR into the generated article, I probably should (it's easier for the human to delete something that isn't interesting than it is to include it). So I think I'd only exclude stuff that it's hard to imagine as ever being relevant to a Wikipedia article (like the procedures/exemptions). In the same vein, I can't see what I might usefully do with the "Type" field of the References. At the moment, I am generating cite-web templates if there is a URL present, or cite-book if there isn't. I have no idea what I would do with "tourism" vs "written" type. I am guessing that the NSW HR means "written" to be "reliable source" and "tourism" to mean "puffery" in Wikipedia-speak, but whether that's a basis for inclusion, I dunno. Kerry (talk) 22:41, 23 March 2018 (UTC)
 * Yes Kerry, the Infobox building, Infobox bridge, Infobox dam, etc., supports relief maps and generally I've been using the NSW map for non-greater Sydney locations and the greater Sydney map for locations that fall within that map's scope. I use the parameter |map_type=Australia New South Wales; |map_relief=yes; etc. Please refer to Denison Bridge as an example; or Sydney Opera House as a variant where options exist to select map location in Sydney, in NSW, and in Australia. Also, I'm not too sure what you did with the QHR coords, yet is there an Australian MOS preference for decimal over dms coordinates? I'm an old skool person myself, so I prefer dms; but, if not already in place, we should gain consensus on which is used for all Australian geographic features. It's a bit hard to assess the "tourism" source, as the links are dead and my search of Bathurst Regional Council website yielded no tangible results for 'Denison'. The written sources appear, on face value, to be reliable. Rangasyd (talk) 03:41, 24 March 2018 (UTC)
 * Relief map - done. Most of the weekend was spent on the NSW wikifier. The article in User:Kerry Raymond/sandbox is now using the NSW wikifier (instead of the Qld one) and the architectural elements wikifier (as before). It is doing an OK job at wikifying the article but there are a couple of errors, the link to John Russell is to an artist rather than to an engineer as you might expect (because there is no article for the mentioned engineer) and a mention of Berry Park is linking to the town Berry. Wikifying is an imperfect task. It is looking at stream of words comparing them against a set of article titles (or aliases for them) and trying to see what it can link. It works with "words"; it does not understand their meanings. Mostly the wikifier is working with proper names (of places, of people, of organisations). The "Berry" example is a typical example of what can go wrong, single words tend to get more wikilinking errors. The wikifier is case-sensitive so, it would not have wikilinked "They ate a berry pie" but it would cheerfully wikilink "Berry pies are served in the cafe" to the town of Berry article. There about about 45,000 articles in the NSW wikifier (but probably it won't link to most of them ever). As I say, the wikifier will always be inherently imperfect, *but* we can improve its performance by altering its rules. As I have it set up, it will wikilink to any exact article title match OR an alias. The current set of aliases are sort-of machine/hand generated. Any article disambiguated with ", New South Wales" or "(New South Wales)" can be matched against its undisambiguated name, e.g. "Berry, New South Wales" will link to that town as will plain "Berry". We can add articles and/or aliases to the wikifiers list (which was initially constructed from the set of all articles directly or indirectly in the Category New South Wales). This will cause it to wikilink more things (rightly or wrongly). If it is constantly adding a wikilink we don't want, we can take away the article or alias to stop that behaviour (e.g. we could remove Berry as an alias for that town). I had to stop "Howard" as an alias for the town of Howardin Queensland, because it seemed that almost every QHR article mentioned a man called Howard Something which was wrongly linked to the town. It was less work for me to add the wiklink manually when the word Howard really refers to the town than to remove it most of the time. So, as we progress with rolling out the NSW SHR articles, we can modify the wikifier if it seems to be frequently not linking something it should or frequently linking something it shouldn't. If problems occurs onlyoccasionally, we fix it by hand in the individual articles. The aim is to get the wikifier good enough to minimise human work of fixing it when it gets it wrong. I'll now try to add in some other material found in the NSW SHR entry, e.g. condition. Kerry (talk) 05:17, 26 March 2018 (UTC)
 * I'm still chugging away with the generator (Easter, Commonwealth Games, and other things having delayed things a bit). As we speak, I am making the first attempt to generate an article for the approx 2000 NSW State Heritage Register entries; the generator is up to SHRNo 1908 as I write this. Now this is not saying the job is done, but (fingers crossed) it will be the first time the generator has been able to generate all the articles without hitting a fatal error (previous attempts found plenty of fatal errors, caused either by a bug in my code or my inadequate grasp of just how bizarre some of the contents of the NSW SHR entries actually is!). There are still some sections in the NSW SHR entries that I am not including into the article, but I think they should not be too difficult (famous last words). The harder problems remain the citations in the entries and the photos. I think I may be able to make some progress with the citations (at least for ones that use the common formats). Photos are likely to remain in the "too hard basket" (not all of them are covered by the CC-BY licensing due to being taken by 3rd parties). Kerry (talk) 15:08, 6 April 2018 (UTC)
 * However, one thing that has become apparent to me is that there seems to be some confusion between the New South Wales State Heritage Register (about 2000 sites) and the NSW heritage database which contains many thousands of entries including the State Heritage Register sites but also including many others, some of which are local government entries (or other organisations) and some of which appear to be heritage application/assessments since the entry does not list any heritage listing by any organisation. Glover cottages is an example of the problem. It is a City of Sydney local government heritage listing and not a State Heritage Register listing, but the infobox refers to the designation of New South Wales Heritage Register (which isn't well-defined), there is a citation using Template:Cite NSW SHR whose name and some (but not all) of the documentation suggests it is for State Heritage Register sites, and a Category:Houses listed on the New South Wales State Heritage Register (which is quite definitely incorrect). I think we need to unscramble this omelette. What I think we need to do is to rename the existing template to be something like Template:Cite NSW HD which enables us to cite anything in the NSW Heritage Database and does not imply any heritage listing (so slightly alter what the template emits and the documentation). Then create a new Template:Cite NSW SHR which is only used for State Heritage Register sites and includes the State Heritage Register number as a field (in addition to the the Heritage Database number). This template could include Category:New South Wales State Heritage Register as part of it to automatically apply the category. I think we need to go back to the designation registration and replace the ill-defined New South Wales Heritage Register with the precise New South Wales State Heritage Register so the infobox box designation is precise. I suspect the reason that this confusion came about is because the NSW designation and template is copying from the Queensland equivalents. But in Queensland, the Qld Govt's heritage database only records state heritage listings. The local governments etc record their local heritage in various ways, but not in the state heritage database. Thus there is no need to distinguish between the Qld Heritage Register (and no need for the term Queensland State Heritage Register) and the Qld Heritage Database as the state database only contains the state register. This is not the case in NSW. So, I think we need to sort this out, as I need to know precisely what designation, what template and what category I generate for NSW State Heritage Register entries. At some later time, we could go back and created designations/templates/categories for the various NSW LGAs (and any other heritage registers) and tidy up the articles like Glover's Cottage that whose heritage register status is a bit misleading with the current template, but none of that affect my article generation so it's not on the critical path for me for generating NSW SHR articles. Kerry (talk) 15:08, 6 April 2018 (UTC)
 * Template:Cite NSW HD has been established! YAY!! Sorry for the delay. Also requested the addition of a new designation for items of non-NSWSHR significance at WikiProject Historic sites. The ball is rolling. Rangasyd (talk) 12:47, 15 May 2018 (UTC)
 * Great! Now with Template:Cite NSW SHR, can you add an optional parameter for the SHR Number, which we don't currently have a way to output in the citation, e.g. 01665 for Denison Bridge. I think it has to be optional at this point because of any existing uses of the template. I imagine over time we can go through the existing uses of the Cite NSW SHR can be inspected and either have their SHR Number added or be converted to use the Cite NSW HD as appropriate. Once that's done, we could change the template to make SHR Number mandatory. Once I know what that new field is and where it goes in the template use, I can generate that. I still have to pull in my new "try to spot the person name" code into the generator (it's only in a test harness at the moment). But with both of these things done, I think we can then start the process of rolling out some articles and tweak the generator (including the wikifier) as we go. Getting close I think! Kerry (talk) 13:24, 15 May 2018 (UTC)
 * Done. Yeah, but how do I make the optional identifier "appear" at Template:Cite NSW SHR and/or Template:Cite NSW SHR/doc? Rangasyd (talk) 14:00, 15 May 2018 (UTC)
 * Not sure what you mean by "appear", but generally with templates, if your changes to a template don't seem to be working, it often means you have to purge the template to force Wikipedia to use the new version of the template. I purged the template and changed the generator to produce an "hr" field and you can see the results at User:Kerry Raymond/sandbox for Elizabeth Farm (which is entry number 1, but as the NSW SHR presents it as 00001, I've been consistent with that. So if you want to use that article in the sandbox to do any fine-tuning of the template representation, go ahead but remember you will probably have to change the template, purge the template and then purge the sandbox to force it to re-render using the updated template. Maybe you won't need all those purges, but I find you need to purge a lot when tinkering with templates. Next step for me is to get the new and improve author detection code into the generator. Also I probably need to tinker with the CC-attribution template at the bottom of the article to be more consistent with the style it's cited, but that can wait until you say you are happy with the cite NSW SHR presentation and then I'll tinker with the CC-attribute template to be consistent with what you do. Kerry (talk) 06:45, 16 May 2018 (UTC)

Now something I haven't talked about until now (because I completely forgot about it) is the need to add lists of the various heritage places into their suburb/locality article. There are a number of benefits to doing this before rolling out the articles themselves. Firstly, it means that when you do create the articles, they will not be orphans (as they will always be linked from their suburb). Secondly, it will be the time to decide whether or not the titles for the article are suitable. Now the good news is that I have another generator to make these lists. The bad news is that, unless told otherwise, the generator assumes that the article title and the lede title (the bold thing in the first sentence) will be the same as the official name in the heritage register. Unfortunately many names in the heritage register are entirely unsuitable for Wikipedia articles. To illustrate this, take a look at the list generated for Millers Point (note, this is one of the worst ones -- most aren't so bad). As you will see, it is awash with problem names, e.g. simple names like Shop, Building, Terraces (many of which are repeated even in this suburb) which are too common to use as article titles. Also even when the names seem more reasonable like Lord Nelson Hotel, we discover to our horror that this existing article is for a hotel in Canada, so we have false positive names to disambiguate. We also have false negative names like Garrison Anglican Church Precinct which is a redlink but alas there is an article about that church already called Garrison Church (Sydney). So, what do we do about this mess? Well, for the too-common names, you need to read the NSW SHR entry and see if you can see anything more specific that might make a better article title. For the first Shops at 1, 3, 5, 7 Argyle Place, maybe you might call them Federation-era Shops, Millers Point or given the number of heritage entries called Shops in this suburb, you might need to all it Federation-era Shops, Argyle Place, Millers Point. If you can't think of anything to call them, the option of last resort is just to use the address "1, 3, 5, 7 Argyle Place" as an article (I reached that desperation point with the QLD Heritage Register a few times). For the false positives like Lord Nelson Hotel, the solution is the usual Wikipedia disambiguation, e.g. Lord Nelson Hotel, Millers Point or Lord Nelson Hotel, Sydney to disambiguate. For the false negatives like Garrison Anglican Church Precint, you need to read the suburb article for any links to things that might potentially be heritage listed. In this case the Millers Point, New South Wales article already mentions Garrison Church which resolve that one. This exercise will also reveal that the Fort Street School mentioned in the Millers Point article is linked to the wrong school (it's in Parramatta) and even when you find the Fort Street Public School (which is in Millers Point), it doesn't appear in our list of heritage properties. Further investigation reveals that it is not individually heritage listed but is in fact part of the listing Millers Point & Dawes Point Village Precinct. While all of this is somewhat painful, it is an important to know these things before you roll out the articles in that suburb. (And as I say, Millers Point is one of the worst suburbs for this example, most of the suburbs are pretty straightforward). If it is not obvious, the list of heritage places that I generate for each suburb is sorted alphabetically by street name, then street number, then article title. This works fine if the NSW SHR entry has a straightforward addresses like "5 Main Street, Smallville" but will break on some of the weird stuff they put in the street addresses. If there's some weird stuff in the street address, you may want to reorder that entry in the list into a more appropriate spot and/or reorganise the way the street address information is presented. Millers Point doesn't have weird addresses but something like "Turn left off Little Creek Road, Smallville", will think the street name is "Turn ..." and sort it accordingly. The generator assumes the first uppercase letter is the start of the street name and the first number found before that is assumed to be the street number, so "right turn at 12.3 km along Little Creek Road, Smallville" will sort as if it had said "12 Little Creek Road, Smallville". These assumptions mostly produce an acceptable result in practice, but like anything produced by a generator, may need some tweaking based on human interpretation of what was being said. I would strong encourage the list of heritage places to be rolled out before any actual articles as it helps finalise the article titles. Note, the generator is happy to be instructed as to a better article name and to a better lede name (they can be different), just tell me what changes you want after rolling out the lists. (Aside, the citations in these lists do not currently include the file numbers, this is just for speed in development/testing as extracting the file numbers involves a lot of extra webscraping which massively slows down my development cycle) but I can do final versions with the file numbers. Kerry (talk) 03:41, 17 May 2018 (UTC)
 * Makes sense. What would you like from me? The Drover&#39;s Wife (talk) 03:57, 17 May 2018 (UTC)
 * One thing I'm not clear on looking at Millers Point - are these actually all separately listed? The house articles all seem to contain relatively little information and all mention that they're in the Millers Point Conservation Area. We probably could get a stub out of these reworking everything that's there, but there's going to be no real possibility for expansion on these nameless houses. The Drover&#39;s Wife (talk) 04:06, 17 May 2018 (UTC)
 * Yes, they appear to be all separately listed in the NSW SHR, but I think your comment about the content being a bit thin is valid. It may be that merging them into a single Millers Point Conversation Area is the way to go. But my sense is that we may do better to start work initially on some of the not-so-heritage-rich LGAs where the roll-out of the listings will be much more straightforward and have our learning curve on them. If we leave the awful ones (and Millers Point seems to be a highpoint for problems) until later in the process, we will have built up a lot of experience before tackling them. My preferred sequence for the roll out is:
 * roll out the heritage listings into the suburb/locality articles one LGA at a time, starting with the easier ones
 * roll out the articles one LGA at a time, starting with the smaller ones
 * That approach let's us build up experience in ourselves and the tweaks needed in the generators, before moving into the heritage-rich areas in inner Sydney, Parramatta, etc. I was thinking in terms of everyone who is interested to be involved to be working on one LGA at a time (to avoid tripping over each other if we all just pick articles at random). I'll probably break up my spreadsheet that tracks work in progress into smaller spreadsheets, one per LGA, which I will give to the person doing that LGA. That way they can update the spreadsheet on any changes to article title, lede name, and various status fields (done nothing, uploaded as new article, existing article no action, merged with existing article, uploaded as 2nd article) and send that back to me from time to time, so that I can update the master spreadsheet so we know where we are at.
 * Aside, should we be moving this conversation to some more appropriate place, such as WikiProject Australia or WikiProject New South Wales, particularly if we hope to recruit more people. Kerry (talk) 06:48, 17 May 2018 (UTC)
 * Conversation moved from my talk page to this page for ongoing maintenance and input. Happy and constructive editing and input from others, we hope! There has also been some offline email contact between Kerry and me, if you want more detail, otherwise the majority is covered above. A big credit to Kerry for establishing her article generator for the 2,000+ state heritage register items and then the many more thousands that are on local government registers, that do not meet state, national or international heritage guidelines. Rangasyd (talk) 09:28, 17 May 2018 (UTC)
 * The Sydney LGA is not the best place to start. It is full of heritage items! Re The Lord Nelson Hotel, according to its website, the title is The Lord Nelson Brewery Hotel. Re Millers Point, that's a bit more challenging. We could simply do an article covering the Dawes Point and Millers Point conservation areas (and here), which is very topical     or we could group them by street; e.g.:
 * Terraces and shops in Argyle Place, Millers Point
 * Terraces and shops in Bettington Street, Millers Point
 * Terraces and shops in Dalgety Terrace, Millers Point
 * Terraces and shops in High Street, Millers Point
 * Terraces and shops in Kent Street, Millers Point
 * Terraces and shops in Lower Fort Street, Millers Point
 * Terraces and cottages in Merriman Street, Millers Point
 * Terraces and shops in Trinity Avenue, Millers Point and
 * Terraces and shops in Windmill Street, Millers Point.
 * Thoughts on next steps? Rangasyd (talk) 11:32, 17 May 2018 (UTC)

Specific example feedback
I decided to do a real-life trial. I picked Young, New South Wales and added the generated list of NSW SHR sites to it. I then rolled-out the 3 generated articles for it, Young railway station, City Bank building and Blackguard Gully. It is probably looking at the diff of my first edit and my last edit on each to see the amount of copyediting I did. Was it successful? Well, adding the generated list to the Young article was good. The Blackguard Gully produced a reasonable article for relatively little effort (there were two cited sources not listed in the NSW SHR, one I tracked down and the other I could not track down -- see the Talk page). Young railway station was completely lacking a history and the description was minimal, but it turns out that this is the fault of the NSW SHR. But there is content on the history etc in the *local* heritage register, so there is potential flesh it out. The City Bank article is totally minimal, again the fault of the NSW SHR, but again there is content in the local heritage register which could flesh it out. So, I think I need to develop another generator for pulling content out of local heritage register, not to produce a full article but do the wikifying and citation processing and an attribution so it's easy for us to pull in content from these other entries in the NSW Heritage Database, as I think we are likely to need it! So overall I would say it was successful in terms of the generator but let down by the NSW SHR entries being rather sparse. Kerry (talk) 03:51, 18 May 2018 (UTC)
 * Also sounds like an excellent idea. Blackguard Gully is a great example of why this project is so important. The Drover&#39;s Wife (talk) 13:05, 18 May 2018 (UTC)
 * Well done Kerry. Several hours later, I've edited just one of those articles, Young railway station and added some additional content, refs, etc. A few things in no set order:
 * Can we please make the coordinates dms and not decimal?
 * I much prefer decimal coords myself and that is what the NSW SHR uses. But the way you see coords displayed is a [|preference you can manage for your own display purposes]. But if there is community consenus for one or the other, I probably can convert them to DMS. Kerry (talk) 02:44, 19 May 2018 (UTC)
 * Can we please add Use Australian English and Use dmy dates above infobox?
 * Yes, this can be done. Kerry (talk) 02:44, 19 May 2018 (UTC)
 * Done. Kerry (talk) 21:33, 19 May 2018 (UTC)
 * Is there a reason why not all the fields in Infobox historic site are not copied?

The fields that are included for the infobox are based on the "most used" as in blank template plus whatever else I needed for the info in the heritage register itself. While I can certainly add all the other fields with blank values, I note people do get grumpy at you for having templates with lots of empty fields. I'm open to community consensus as to what to include.
 * Is it possible to make all the first letters for each field in the infobox Capital and not capital :-)
 * Yes it's possible technically, but there's MOS issues about not overcapitalising. Personally I don't care, but some in the community do care. Kerry (talk) 02:44, 19 May 2018 (UTC)
 * The lede and the infobox feels a little "clumsy" when we list suburb, local government area, New South Wales, Australia. Is the LGA really necessary?
 * It is necessary in Qld as suburb names are reused in different LGAs. Not sure if this is the case in NSW. It's easy technically to leave it out. But I think it is helpful if general as it assists the reader have a better sense of where it is particularly if it's in a suburb/locality rather than a larger town (which is likely to be better known). If I say it is Fishery Falls, Queensland, you are gonna say "huh, where in Qld, it's a big place?", but if I say Fishery Falls, Cairns Region, Queensland, it's a lot more likely you will now have a rough idea of where it is. Kerry (talk) 02:59, 19 May 2018 (UTC)  Kerry (talk) 02:44, 19 May 2018 (UTC)
 * Can we add a ==See also== section with New South Wales
 * No problem to do this. Kerry (talk) 02:44, 19 May 2018 (UTC)
 * Done. Kerry (talk) 21:40, 19 May 2018 (UTC)
 * Should we wikify heritage-listed as heritage-listed at lede; and should it be an heritage-listed, as opposed to a heritage-listed?
 * Technically it's easy to do, but I got "sea of blue" complaints about wikilinking it in the QHR articles, so it's a consensus question. The a/an choice seems to depend on your personal dialect of English (how much you pronounce the "h") and/or where/when you went to school. Not sure if the MOS has anything to say on it. For my pronunciation and schooling, it's "a heritage-listed". Open to community consensus. Technically it is easy to do either way before "heritage-listed" but not so easy for the generator to figure it before arbitrary words in the lede relating to current and former types of the site. Kerry (talk) 02:44, 19 May 2018 (UTC)
 * And if a measurement unit is displayed (km, mi, kg, m, ft, etc., is there an automated way to insert convert
 * Probably for single units, I need to experiment.
 * I have implemented it for km and miles. But it's a case of "be careful what you wish for". For example, if I use it to generate the Blackguard Gully entry, it produces:

The area known as 'Burrangong Goldfields' covered an area of 20 mi (32 km) by 10 mi (16.5 km). because the original text contained a manual conversion already: The area known as 'Burrangong Goldfields' covered an area of 20 miles (32 km) by 10 miles (16.5 km). But on the principle of "it's easier to delete something wrong than add something", it's probably still worth doing notwithstanding the possible over-enthuasiasm in cases like this. Now to make this all work, I do need to know the weird and wonderful array of units used in the NSW SHR and how they are expressed. For example I will match any of "km" "kms" "kilometre" "kilometres" as being alternative ways to say that unit, but for miles, I only allow "mile" and "miles" as I am not sure if "mi" or "mis" is likely to occur in the NSW SHR. And when it comes to metres, it is a little more risky. The letters "km" don't occur that much in normal English with any other meaning, but when you start matching single letters like "m" for metres, the chances of converting an "m" that doesn't mean metres are much increased. For example, "the railways used their new Class 34m locomotives on this route" is not saying "34 metres" but the generator will change it to ""the railways used their new Class 34 m locomotives on this route" ". As always with the generator, the smarter we make it comes with the downside of having the human user to be smart enough to notice when it gets it wrong and work out what was really intended, which is usually best done by looking back at the NSW SHR entry and seeing what the original said. So to do other units, I would appreciate if people could point out examples of their use as I need to be as careful as I can be in the replacements to reduce the likelihood of undesired conversion as illustrated above. Kerry (talk) 23:50, 19 May 2018 (UTC)
 * All the other changes that I made were subjective based on the additional source material. In the infobox the following field are somewhat limiting, especially where there is no end_label
 * |beginning_label
 * |end_label
 * Now to review the others!! Rangasyd (talk) 13:35, 18 May 2018 (UTC)
 * The infobox only has a beginning_label, there is no end_label? Or are you saying the infobox should have an end_label in its definition? Kerry (talk) 02:53, 19 May 2018 (UTC)
 * The Infobox historic site could benefit from an end_label in its definition. While you're there, is it worth looking at Infobox building and seeing if we want to add any additional fields to Infobox historic site? The historic site template appears to be a little light on. I've reviewed Blackguard Gully and City Bank building and made some minor changes. Rangasyd (talk) 16:29, 19 May 2018 (UTC)
 * If you want to change Infobox historic site, I'd suggest discussing it on its talk page. I just use the template; I've had no involvement in creating it. I can't see why there should be an objection to having an end_label with similar semantics to the beginning_label. I am less sure about loading it up with fields from infobox building; I suspect the preferred solution might be to create a module structure which enables two infoboxes to be combined as one (but I confess I have never done this, so I am vague about how to do it). Kerry (talk) 04:37, 21 May 2018 (UTC)
 * I'm also noticing that the RNE has better entries than the SRHP for some of these places - I know nothing about the technical side of things, but would it be possible to be able to pull this stuff in the same way as the local one you're working on? The Drover&#39;s Wife (talk) 23:17, 19 May 2018 (UTC)
 * Yes, in theory, but not as easily as the NSW local heritage register entries. The NSW local heritage entries are in the same web format as the NSW SHR entries(give or take a few details), so I can more-or-less reuse the webscraper that I already have (so technically easier) plus the NSW local heritage register seem to use the same criteria structure as the NSW SHR, so we can probably re-use the significance/criteria text (good content). The RNE is a completely different web page structure so needs a completely new webscraper development (more work technically), plus I don't know if the significance/criteria used in the RNE is sufficiently similar to the NSW SHR to be able to reuse that content (dubious content). We should be able to use the RNE as a substitute for a missing history/description info though. Figuring out how to exploit the RNE is fundamental to doing any generator for the other state/territory heritage registers as they either don't publish any narrative content on the web (e.g. TAS and SA) or do so but it's copyright (e.g. VIC). Kerry (talk) 04:37, 21 May 2018 (UTC)
 * If I can plead anything, please do not use the RNE as a primary source. The RNE was established in 1976 and abolished in 2007, replaced by the CHL and the ANHL. The problem with the RNE is that the criteria was so loose anything could be listed, most notably during the mid-1980s when the Hawke and Keating governments went wild. Now, in order to get listed on the CHL or the ANHL, the place needs to be first nominated by a state/territory government. Unfortunately (or fortunately) the AHD still lists places on the RNE. But please do not use it as a reason for creating articles. If I can suggest a work order, we should complete all state and territory heritage listed laces before completing any places listed on the CHL/ANHL. In face, I think all the places listed on the ANHL have been done (but not certain). I know that not all the places listed on the CHL have Wikipedia articles. Rangasyd (talk) 05:59, 21 May 2018 (UTC)
 * I wasn't talking about a separate project for the RNE - I've never been keen on that, largely because (while I'm not sure I agree with you about notability) so many of the RNE entries are extremely brief that it'd wind up being an exercise in mass stub generation. My comment here was about the possibility of using RNE text to supplement the SRHP text where the former is indeed superior (as happens often enough with the SHRP's tinier articles). This question also arises with AHDB more broadly in relation to the other states, as Kerry said - e.g. I've used content from statements on the denial of a ANHL listing to build articles for sites on the Victorian Heritage Register, due to the copyrighting of the text of VHR entries. It's a bit of a moot point here, though, if the AHDB generator is too much of a hassle to build in time to be useful on the SRHP rollout - just another thing to chalk up on the long-term wishlist. The Drover&#39;s Wife (talk) 06:14, 21 May 2018 (UTC)
 * Just to inject a note of reality into this iscussion. It's going to take a lot of time to roll out the NSW SHR articles, far longer than the time it will take me to build these additional generators. If we don't feel that some NSW SHR articles are immediately viable due to lack of NSW SHR content, we can decide not to roll them out right now or roll them out with incomplete tags, but if we can see some promising content in the LHR or RNE, we can flag that fact in the master spreadsheet that manages the project and come back to those articles when the additional generators are ready. I have no intention that the NSW SHR generator will automatically incorporate any/all content it can find in some vaguely-similarly-named LHR or RNE entry (a rather high-risk strategy!). No, a person (like you) will decide if the content in one of those other sources is useful, then my tool will do its usual job of wikifying, adding citations, generating attributions, etc, and the person shall incorporate as much of it as they think appropriate and do the usual copyediting. Think of them as "mini-generators" to provide supplementary article content as they will probably not be full generators. I think we are all agreed the RNE is a defunct register and should not be the sole basis for a "heritage" article, but we may find situations when it can provide some history or description for a site listed in a "real" heritage register. This will particularly be the case for Victoria and other states where the heritage registers are not CC-licensed. I suspect for the NSW SHR articles, we will get better value from NSW local heritage entries than from the RNE, so I will be prioritising the LHR "mini-generator" over the RNE mini-generator for both that reason and because the LHR one should be much easier to write. Meanwhile I will go back to trying to finish the conversion of units in the NSW SHR generator, a task which (as always, sounds much easier to do than it is in practice), as nothing is going to happen until I've done that. Kerry (talk) 07:20, 21 May 2018 (UTC)

Categorisation
Should Category:New South Wales State Heritage Register be fully diffused, except for the header/main article/s with all register entry articles going into relevant register category subcategories, of which several already exist? Aoziwe (talk) 12:42, 19 May 2018 (UTC)
 * Once complete, we will end up with ~2000 entries in the category, without breaking it down into sub-categories. I'm ambivalent, either way. I started the sub-categories, but happy to take advice on what we think is best. And if we decide to sub-categorise, how is it best done? By type of place listed (e.g. railway, building, natural place), by year listed, by region in NSW, etc.? Rangasyd (talk) 16:22, 19 May 2018 (UTC)
 * From what I know from the Queensland Heritage Register, the range of types of heritage-listed things is much wider than anyone could imagine. Yes, there are lots of homesteads, churches, schools, railway stations etc. But there are cane lifts and Chinese pig roasting pits and Cornish boilers and all kinds of other whacky stuff. So I think full diffusion is not likely to work well here as it may produce a lot of single-entry categories for all the whacky stuff. Kerry (talk) 00:39, 20 May 2018 (UTC)
 * Indeed, I would not subcategorise the heritage register category at all. IMHO, we massively "over-categorise" in Wikipedia, which is why we end up with categories like female 19th century authors, which is essentially the intersection of 3 separate category trees, by sex, by time frame, by occupation. But because Wikipedia does not directly give the user a tool to say "show me all the female 19th century authors", we generate zillions of these ridiculous over-blown categories as a poor man's alternative. (Tip: there is a tool called petscan at the WMF labs website that will do category intereection but it's not provided in a way that is useable by readers or even findable by most editors). In an ideal Wikipedia, the Young railway station would be in three categories: railway station (in the type category), Young (in the place category) & NSW SHR (in the heritage category), as each addresses a separate concept. But we live in a sub-optimal Wikipedia, so we have to have some of these cross-categorisations to support the likely user needs for category intersection. But nonetheless some cross-categorisations are better than others. Saying that Young railway station is in the category "Railways stations in NSW" is not that bad because these are both "fundamental" characteristics of the Young railway station. It is a railway station, it is in NSW, it's not going jump and move to Mexico, nor transform into a species of butterfly (it might be *used* for a different purpose, e.g. museum, but it remains in form a railway station -- if you look at the museum, you'd say "it looks like a railway station"). But a heritage-listing is not a fundamental characteristic. The Young railway station was not always heritage-listed and at some futre point a decision might be made to delist it. Doing cross-categorisation of fundamental characteristics with non-fundamental ones is therefore bad modelling, therefore I don't like subcategorisation of any HR categories with fundamental properties like being a railway station. But it would be acceptable IMHO to have subcategories the heritage register category by characteristics of the heritage listing. For example, in Queensland, there are 3 flavours of listing (state heritage, archeological site, protected place) which are things that could legitimately be used for sub-categorisation as they are specific characteristics of heritage listings. I did not ever create subcategories for Category:Queensland Heritage Register for these reasons, although I see someone has added one with 4 entries (sigh). Kerry (talk) 00:39, 20 May 2018 (UTC)
 * I think we should go with no diffusion, and hence no subcats as per current cat structure. I agree with Kerry.  Unstable and-or non fundamental characteristics for categorisation is problematic.  If there was some broad categorisation related to the heritage register itself then that is fundamental to the heritage listing, as per Queensland for example, then we could use that.
 * We could add to the category lead something like:

''To find articles about a particular type of heritage listed feature, use the "incategory" and "intitle" search keys in a wiki page-search. For example, to find all articles about lighthouses listed in the heritage register type the following into the wiki page-search field:'' incategory:"New South Wales State Heritage Register" intitle:"lighthouse"


 * For example, Category:Bridges listed on the New South Wales State Heritage Register‎ would be essentially replaced with incategory:"New South Wales State Heritage Register" intitle:"bridge". This search currently returns no results because such articles have been fully diffused.  Aoziwe (talk) 11:52, 20 May 2018 (UTC)


 * Or we could have in appropriate places, for example, the category lead and at the end of the New South Wales State Heritage Register‎ article:

Look for New South Wales State heritage registered features of type:

Aoziwe (talk) 12:41, 20 May 2018 (UTC)
 * The search box above now works with subcategories. So, for example, replacing <..> with bridge will return all registered cat bridge articles.  Aoziwe (talk) 10:43, 21 May 2018 (UTC)
 * Despite my ranting (which is why I labelled it as such), I can live with having subcats of the NSW SHR category where there are a lot of things of that type and there will certainly be lots of NSW heritage-listed churches and railway stations etc. I just don't want to do full diffusion for every weird and wonderful type of thing that's lurking in the heritage register. Full diffusion only really makes sense where there are a fixed finite set of choices over which the diffusion can be done. Kerry (talk) 07:37, 21 May 2018 (UTC)
 * Okay by me. We just need to document that approach somewhere, in the cat lead perhaps ?  Aoziwe (talk) 10:43, 21 May 2018 (UTC)
 * I was confused when I saw removing the Churches listed on the New South Wales Heritage Register category from articles, so I found the discussion here. Was this the consensus? Are the other subcats Bridges..., Government buildings..., Houses... and Rail infrastructure... also being removed? To me it seems counterintuitive to merge subcats to one category with over a thousand articles (maybe eventually thousands). Will the explanatory text to use the "incategory" and "intitle" search keys be added to the category? Jack N. Stock (talk) 14:59, 16 November 2018 (UTC)

The discussion just died out it seems. Given that Kerry and I had agreed and no one else disageed, I had assumed the approach was to have sub cats for the bigger groupings and the left over relatively obscure register items would not be diffused? Which was the emerging de facto status anyway? Aoziwe (talk) 04:09, 19 November 2018 (UTC)

Discussion
This project looks like a good topic for discussion at next weekend's Sydney meetup (28 May). If any of the participants can make it along to that meetup, you're very welcome to join in. Bahnfrend (talk) 03:29, 20 May 2018 (UTC)
 * Thanks. Would love to attend but have a prior commitment. Please keep me posted on the outcome. Been working with Kerry on this. Rangasyd (talk) 04:01, 21 May 2018 (UTC)
 * I would have liked to come but I was in Sydney last weekend not the coming one. Such is life. Kerry (talk) 07:40, 21 May 2018 (UTC)

WP:SUBCAT
Rangasyd has recently deleted every sub category for every item in the SHR list. Seems a pointless task, WP:DIFFUSE, was there a need ? Will the 1400+ items in the subcat for entries with text from the SHR be also deleted ? Dave Rave (talk) 20:37, 17 November 2018 (UTC)


 * Kerry, Aoziwe and I were both not keen on the cross-categorisation in this context - she posted a very detailed rationale for why back in May. I'm glad Rangasyd finally got around to doing (what at least I view as) cleaning up the mess. The SHR text subcategory is a hidden one and serves a different purpose. The Drover&#39;s Wife (talk) 20:43, 17 November 2018 (UTC)
 * was that the sub SHR v HD debate or this greater every single subcat idea ? Dave Rave (talk) 21:53, 17 November 2018 (UTC)
 * It seems to me to be a backward step, as this could now be considered an overpopulated category requiring diffusion (as Dave Rave suggested). If we wait a while, some other editor will come along and subcategorize. It also was executed without an easily-located explanation, and it took me a while to find the discussion. Jack N. Stock (talk) 21:25, 17 November 2018 (UTC)
 * i don't mind using a different naming convention, but it was already there, getting rid of it just boosts contrib numbers Dave Rave (talk) 21:54, 17 November 2018 (UTC)
 * It was a lightly-rolled out move predated the content project and was abandoned fairly early in it - the only articles that were categorised in this way that were added in the project (the overwhelming amount of the articles in the category) were ones that somebody uninvolved decided to recategorise in them later. The Drover&#39;s Wife (talk) 23:11, 17 November 2018 (UTC)
 * Well, slap the hand of that "somebody uninvolved" because I'm sure the few people commenting on the earlier categorization discussion WP:OWN the articles and categories on WP. Compare Category:Buildings and structures on the National Register of Historic Places in Florida (picked at random) for another example of categorization. Jack N. Stock (talk) 00:24, 18 November 2018 (UTC)
 * I didn't say that, I was just pointing out that it wasn't already there in the overwhelming amount of cases. There are obviously other ways of doing this, just that the majority of users who've responded on this issue seem to be unpersuaded that it is useful to subdivide it in this particular way. The Drover&#39;s Wife (talk) 11:12, 18 November 2018 (UTC)
 * Three people had a conversation at this obscure WikiProject talk page. That doesn't represent a meaningful majority. This has not been discussed. The fact subcategorisation was not fully implemented is not relevant because WP is a work in progress. Others are going to subcategorize later because it is a WP guideline to WP:DIFFUSE large categories. Review Category:Overpopulated categories and you will see that the Australian state heritage register categories are within the consensus of size for "overpopulated categories". Jack N. Stock (talk) 14:28, 18 November 2018 (UTC)
 * And two other people on the same talk page doesn't override that just because they have very strong opinions. There does not appear to be any hard and fast guideline that large categories have to be diffused if it doesn't make sense to do so. The Drover&#39;s Wife (talk) 02:15, 19 November 2018 (UTC)

OK, that discussion seems unlikely to get anywhere. Another idea: how about using the same categories as the NSW State Heritage Inventory? It should be mentioned that the New South Wales State Heritage Register is IRL a subcategory of the New South Wales Heritage Database (though not on WP). - Jack N. Stock (talk) 03:14, 19 November 2018 (UTC)
 * Section 1 - contains Aboriginal Places declared by the Minister for the Environment under the National Parks and Wildlife Act. This information is provided by the Heritage Division.
 * Section 2 - contains heritage items listed by the Heritage Council of NSW under the NSW Heritage Act. This includes listing on the State Heritage Register, an Interim Heritage Order or protected under section 136 of the NSW Heritage Act. This information is provided by the Heritage Division.
 * Section 3 - contains items listed by local councils on Local Environmental Plans under the Environmental Planning and Assessment Act, 1979 and State government agencies under s.170 of the Heritage Act. This information is provided by local councils and State government agencies.


 * I think it already does - it definitely shouldn't contain any from #3 and I don't think it contains any from #1 - neither of these are actually on the SHR. The Drover&#39;s Wife (talk) 08:11, 19 November 2018 (UTC)
 * I can't win. I'm THAT someone who started the earlier sub-categorisation when I commenced a small project to create articles for all NSW railway stations listed on the NSWSHR. Kerry then became involved and autogenerated articles for ALL items on the NSWSHR, following a similar initiative in QLD. This NSWSHR autogeneration didn't create any sub-categories. There was a discussion between Kerry, The Drover&#39;s Wife and me that the sub-categories be deleted. I was opposed to the removal. Nevertheless,, following feedback and consensus, it was decided to remove. Given that I created the sub-categories I felt a moral obligation to remove. Now I'm told that the removal of the sub-categories was not ideal. My view still stands, with 1600+ items on the register, it appears unmanageable and overpopulated as a category. We should establish a standard for Australia along the lines of that established for the NRHP; and/or the list could be by LGA, e.g., Category:New South Wales State Heritage Register sites in ABC local government area. Oh, and BTW, I don't agree with using the inventory. This contains places of lesser importance and will confuse the lay observer. Rangasyd (talk) 11:07, 19 November 2018 (UTC)
 * Looking back at the previous discussion, it reads like the consensus was to keep the subcats, although Kerry had misgivings and seemed reluctant to agree. The Drover&#39;s Wife says otherwise, and maybe this reflects a conversation at the Sydney meetup on 28 May. somebody uninvolved (talk) 10:47, 20 November 2018 (UTC)
 * "Consensus to keep the subcats" is an incredibly innovative description of a discussion in which literally no one supported keeping the subcats. I can't imagine any relevant conversations were had at a meetup that no one from this WikiProject went to. The Drover&#39;s Wife (talk) 10:52, 20 November 2018 (UTC)
 * Read the end of the "Categorisation" sub-section, starting from "Despite my ranting..." Jack N. Stock (talk) 11:04, 20 November 2018 (UTC)
 * ...which isn't actually "consensus to keep the subcats". The extremely aggressive tone and determination to ignore the opinions of all other editors because you're so hellbent on getting your way exactly how you want it, everyone else be damned, is getting very tiring, and is making an actual consensus outcome of this discussion impossible. The Drover&#39;s Wife (talk) 11:10, 20 November 2018 (UTC)
 * I raised the point, so count me, he stood up, so that's two. Everyone else said go on account of some conversation that no-one else knows about, but we don't care about subcat or diffuse, and as for tone, Drover, really ? isn't it always your way and be damned ? Dave Rave (talk) 05:32, 21 November 2018 (UTC)

I am not really involved in this particular project but I was I think involved in the earlier discussion, and I too am confused. Aoziwe (talk) 11:20, 20 November 2018 (UTC)
 * That's my position, too. I don't have the strong opinion that I seem to have implied, I was just confused when the subcats were removed. The Drover's Wife was right, there clearly wasn't a consensus to keep the subcats, but I can't see that there was a consensus either way. Jack N. Stock (talk) 11:26, 20 November 2018 (UTC)
 * I think there is a misunderstanding about WP:DIFFUSE. Firstly it doesn't say you must diffuse a category (indeed, it makes the point that there isn't a technical limit to the size), only that you might want to. And its examples are all diffusion on natural subsets (Europe into individual countries, Albums by genre, etc). Being listed on a heritage register doesn't have natural subset of railway station. The natural subtypes of a heritage register might be the basis for heritage listing (aesthetic, technological, association with an historical person or event), year of listing, etc. Because the type of a thing (railway station, church) is orthogonal (independent) to its being heritage-listed, this is not a subset but rather a cross-product. Currently all of these NSWSHR articles are generated with three orthogonal categories, the first being its type (e.g. Railway stations in New South Wales), the second being its location (e.g. Bathurst, New South Wales), and the third being its heritage status (New South Wales State Heritage Register). I note that to be truly orthogonal it might be argued that it should just be "Railway stations" and not "Railway stations in New South Wales" but arguably a railway station is a highly immovable object intrinsically linked to its place and that even if a town's railway station is relocated, we normally denote them as different things in real life and hence in Wikipedia (see Old Townsville railway station vs Townsville railway station, Gympie railway station vs Gympie North railway station). And irrespective of the merits of that argument, it's probably too late to unscramble that particular category's existence in any case. If you start diffusing by cross-product, it rapidly leads to Category:Heritage-listed railway stations in Bathurst designed by Snograss opened in 1922 and closed in 1987. While that is a deliberately stupid example, cross-product categories of that nature proliferate on Wikipedia not because they make sense but because we have not given readers (or more to the point, editors, as studies show readers don't seem to use the category system) any easy way to search across the intersection of categories, perform the closure of a category subtree etc. (Note, we do actually have such tools, Petscan being my tool of choice, but these tools live outside of the Wikipedia ecosystem, so most editors are unaware of them). We have a deficient *implementation* of the category system in Wikipedia and so people have come up with bastard solutions like combining orthogonal categories as a poor man's way to deal with categories with many members. If we had given Wikipedians proper tools to work with categories at the outset, say and a button called "Show me everything in the Bathurst category tree", we would probably have a much simpler cleaner category system, because we would have the tools to do the work instead of a more complex category system. However, complex categories like my silly example above get created all the time in Wikipedia, often driven (it seems to me) because the introduction of eah such category provides an endless supply of unproductive edits to be done by people solely interested in increasing their edit count. Such categories do not help the reader (who rarely look at them) and does not help the editor trying to categorise articles in the first place (which really should be addressing the who/what/when/where/why/how categorisation of a topic as a set of orthogonal issues). Fortunately sanity seems to prevail in some cases. Category:2018 deaths (and the like) which have many thousands of entries are not further "diffused" into 2018 deaths in Bathurst, 2018 deaths of architects in Bathurst, 2018 deaths of male architects in Bathurst, 2018 deaths due to lung cancer of male Russian Orthodox architects in Bathurst, etc. Curiously, they don't even do the one thing that would be justifiable as a diffusion: January 2018 deaths, etc. End of rant (sorry, this stuff was once my day job). Kerry (talk) 22:49, 20 November 2018 (UTC)
 * Personally, I don't think we reached a consensus in the earlier discussion. While I do have very strong views on this stuff and stated my views back then (and again above), I did conclude earlier with "I can live with having subcats of the NSW SHR category where there are a lot of things of that type and there will certainly be lots of NSW heritage-listed churches and railway stations etc". When Rangasyd recently began to remove those subcategories, I assumed that, having given some thought to the matter, he had been persuaded by my arguments that it was the better way, but it seems that was not the case. Talk pages are not a great way to establish consensus. It is not always clear that a discussion has terminated or whether it just petered out which makes it difficult to establish if there was or wan't a consensus and, if there was, what was it. Kerry (talk) 00:05, 21 November 2018 (UTC)
 * Don't think it matters what subset names you use, just a method of how to find all the things of a like nature. Dave Rave (talk) 05:32, 21 November 2018 (UTC)
 * FYI All. Category instersection will work, eg, in the search box enter incategory:"Railway stations" incategory:"Designed by Snodgrass" incategory:"Opened in 1922".  Searching for all NSW heritage railway stations would be incategory:"New South Wales State Heritage Register" intitle:"railway station".  If there are subcategories use deepcat: instead, if you want to get down to four levels of subcat.  Aoziwe (talk) 11:07, 21 November 2018 (UTC)

Template:Cite NSW SHR - making the NSW SHR number mandatory
I did some updates to this template to make the display a bit more lightweight. Previously a lot of information was displayed through the title field, now the heritage register name and number appear in their own separate fields. I updated the documentation to match the template. (the two were somewhat out of alignment in terms of the names of the parameters).

I realised as part of this process that that the State Heritage Register Number parameter (hr) is not mandatory but optional. I think it should be (and indeed have already documented it to be mandatory, but do not actually enforce it in the template definition for fear of breaking some usages) as that forces people to have an SHR number to use this template and therefore discourages them from using it for non-SHR entries in the NSW Heritage Database for which we have a separate template Template:Cite NSW HD (which I have not modified along the same lines, but perhaps should for consistency). In practice the articles which use the Cite NSW SHR template are almost always generated articles (as we set up the template as part of the project) and the generated articles will all provide the heritage register entry parameter, so I don't think we risk breaking anything. Because articles with generated content put themselves into the hidden Category:Articles incorporating text from the New South Wales State Heritage Register, I can use tool like Petscan to find the the small number of articles without any generated content which assert to be in the NSW SHR (currently there are 13 of them) to make sure they use the hr parameter before making it mandatory (i.e. don't break any existing articles with the change to the template). Your thoughts? Kerry (talk) 09:29, 3 November 2018 (UTC)
 * Sounds good? The Drover&#39;s Wife (talk) 09:32, 3 November 2018 (UTC)
 * Agree. It should be mandatory to discourage improper usage; that is, usage of both cite NSW SHR and cite NSW HD should require the heritage register parameter; or any documentation that substitutes usage of the templates, such as designation list, where the value of NSW or New South Wales, etc, is applied. Should this discussion happen at Template talk:Cite NSW SHR and Template talk:Cite NSW HD. Rangasyd (talk) 13:28, 6 November 2018 (UTC)
 * Judging from the history of those templates, all the interested parties are regulars here, as are its regular users (a small but select group!). I will fix the template to throw an error if there is no SHR number provided. I’ll run through the non-generated SHR article to check they are correctly using the hr field first, so no existing article will throw the error. Kerry (talk) 23:48, 6 November 2018 (UTC)


 * is the H part of the id part of the template or part of the shr ? and needs to be a five digit zero filled ? Dave Rave (talk) 00:52, 8 November 2018 (UTC)
 * The "H" has always been produced by the cite NSW SHR and I don't know why that is. The 5-digit-with-zero-fill is the format that is returned from the NSW heritage database search which is what was web-scraped to be the input to the article generator. I don't know if there is any requirement for either of them to be that way. I suspect it's something we can decide to do or not do. It is trivial to change the H in the template. It is not as trivial to change 5-digit-with-zero-fill in the articles but it could be done in AutoWikiBrowser if we thought it was not right. Kerry (talk) 01:30, 8 November 2018 (UTC)
 * I spent well over an hour trying to figure out why the id= wasn't working before noticing that there was two examples and I was editing the first while the second was the problem. So annoying. Dave Rave (talk)
 * are we going to re-implement the folio number part of the template ? Dave Rave (talk) 10:55, 26 November 2018 (UTC)
 * I could live without it. It appears to be a purely internal piece of administrativia. I don't think we have access to the files and nor do our readers. But, OTOH, maybe there is some significance to it, in which case it's harmless to keep it. I guess I am waiting for a clear consensus on all of this before doing it. The actual doing of it probably involves a few steps as it is not possible to change the template and every use of it in a single action, so I probably have to create a new template with the desired structure, use AWB to convert the articles to use the new template, and finally delete the old template. And, then if desired, rename the new template back to match the old template and use AWB to change them back to the old name. (Unless someone can think of a faster way that doesn't leave hundreds of broken articles along the way). Kerry (talk) 02:27, 27 November 2018 (UTC)

The Rocks and Millers Point
Now that we're basically done with all of NSW outside of the Sydney CBD, The Rocks and Millers Point, it seems like these two suburbs are our next target. These are huge, are going to take a lot of work, and are probably going to require a significant number of merges and renamings due to the amount of "House"-esque listings. As per discussion at User talk:Rangasyd, I've dumped the autogenerated suburb lists in a project page at WikiProject Australian historic places/The Rocks and Millers Point, so we can work out what we're doing before putting the lists in article-space due to the amount of likely changes required.

Is everyone okay with working through articles in these two suburbs individually? I feel like it'll be easier to chip away at them as we feel like it rather than trying to work through a set subset of them.

I'd like to be a bit careful about how we deal with the multitude of ambiguously-named places here and make sure that we at least do a quick check - having been the one to do most of the large places with most of the "House"-like names so far, I've found quite a few properties that did actually have names and notable histories. For example, Court House Hotel, Windsor was in the SHR as "Houses" (despite being in the RNE as "Courthouse Hotel") and directly adjoins a bunch of nameless listed properties that I merged as North Street residences. I feel like it's a better outcome to keep individual articles on these sorts of actually-named and more notable properties rather than mashing them together when we merge nameless ones together. The Drover&#39;s Wife (talk) 01:09, 2 November 2018 (UTC)


 * I'm already running into my first complex situation in The Rocks, in Harrington Street. Some local knowledge would be great here.
 * There are four properties in Harrington Street (55-59, 61-65, 67 and 71, all unnamed in the SHR, that seem to have been merged into the same Clocktower development. Three of these would seem like good merge targets, but 55-59 is a remnant of a larger named building with a more interesting history. Not sure what any merged article would be called.
 * Well, having not read any of them (the best way to give advice) I would suggest doing the "the big article" with a short version of 55-59 info and a separate 55-99 article with full info, crosslinked to one another. Kerry (talk) 05:16, 2 November 2018 (UTC)
 * How about Harrington Street heritage buildings (assuming there aren't more HR entries for Harrington Street) Kerry (talk) 05:16, 2 November 2018 (UTC)
 * There is also another unnamed "Terraces" that is now, along with "Evans' Stores", part of the modern Harbour Rocks Hotel. Dunno what to do with these buildings and/or the modern hotel article.
 * I'd do the same thing. If there's enough info to sustain an individual article, do it, and do a short version of it in the other, cross-linked. Kerry (talk) 05:16, 2 November 2018 (UTC)
 * There is also "Shop and Residence" at No. 32, unlike all the others not incorporated in a larger development but with pretty bare-bones SHR content compared to the others. And "Reynolds's Cottages" at No. 28-30.
 * Does anyone see a sensible way to merge any of these? I'd love any ideas but I'm wondering if it mightn't be easier to just create them all separately even if there's some overlap. The Drover&#39;s Wife (talk) 03:42, 2 November 2018 (UTC)
 * There's no policy against overlap that I know of. Kerry (talk) 05:22, 2 November 2018 (UTC)
 * There are other heritage buildings in Harrington Street, sorry - this is why this one is a bit of a puzzle. The Drover&#39;s Wife (talk) 06:25, 2 November 2018 (UTC)
 * Harrington Street heritage buildings (north-east) for 28-52, Harrington Street heritage buildings (north-west) for 55-71, and Harrington Street heritage builings (south-west) for 117 to 127 (assuming you need an aggregate articles for each of these groups). There's a point at which it's time to stop agonising and just do it. If a better names are proposed, we can rename them. Kerry (talk) 06:40, 2 November 2018 (UTC)
 * Okay, I feel like individual articles might be the way to go then - aggregate articles probably too messy/awkward framing. The Drover&#39;s Wife (talk) 07:02, 2 November 2018 (UTC)
 * I did a sweep through all of The Rocks list at WikiProject Australian historic places/The Rocks and Millers Point and tried to put together what I thought were the best likely article names and places where merged articles might be warranted. I generally went with individual articles (with a few exceptions) - The Rocks seems to be so dense with buildings with different histories and reasonably detailed SHR material (unlike the other large SHR places so far) that it mostly seems unworkable to do anything else. Not bothered if anyone wants to change anything/has any better ideas, though will probably start ploughing in in the next day or so. The Drover&#39;s Wife (talk) 08:19, 2 November 2018 (UTC)
 * Wow, Elizabeth Bay took forever! I'm just picking up this conversation now. I agree with tackling Millers Point and The Rocks together. There is much synergy and the boundary lines are unclear. I live about a 30-40 walk from Millers Point/The Rocks. Perhaps you might like to add all your questions together and I can go for a walk about and take photos at the same time and we can resolve any issues. In the interim, it may be worthwhile putting articles where is some confusion into a sandbox before going live. I'm happy to provide practical help where I can. I'm not so much of XXX Street heritage buildings (yyy direction). I think I will need to have a physical look to determine what may be best. I suggest create in a sandbox and move later once there is agreement on the article title. And if you're both okay, I'd prefer to do all churches (and former churches) by using Infobox church, replacing Infobox historic site. Rangasyd (talk) 09:17, 2 November 2018 (UTC)
 * I'm not sure I've got any more questions about The Rocks now that I've thrown together my best guess at a way forward there at WikiProject Australian historic places/The Rocks and Millers Point. The problem is that the practice I've used where there are numerous unnamed buildings in close proximity (like North Street residences in Windsor) doesn't work when in basically every single street in The Rocks there are plenty of other heritage buildings as well, so there's no obvious way to structure a merged article together if one were so inclined. There are numerous cases where multiple buildings have been redeveloped into the same modern development, but "Heritage buildings mashed into the Duty Free Store in The Rocks" isn't really a workable article subject. (I haven't looked at Millers Point - I suspect it might be a bit different there.) Nonetheless - any local light you can shed (and any photos!) would be terrific. I'm not sure what you're suggesting about a sandbox - did you mean the suburb lists or the articles themselves? No objections about the church infobox - I have no attachment to the historic site one, I just dislike manually generating infoboxes and that's what comes in the box. The Drover&#39;s Wife (talk) 09:52, 2 November 2018 (UTC)
 * There is sometimes a reference to 'heritage precincts' that may offer a title, or the adjacent buildings are part of the submissions as 'visual context', though official listings of these are resisted (presumably for the obstructions it presents to development). This may be a longshot, but it has allowed some flexibility in scope in a couple of things I worked on. cygnis insignis 12:47, 2 November 2018 (UTC)

I've just located maps (from the PSMA dataset of Australian Government official boundaries) that define the boundaries of suburbs such as Millers Point, Dawes Point, The Rocks, Barangaroo (and all other Australian suburbs). These maps place some of the properties that are listed by the NSWSHR as being in a different suburb. A good example is to look at properties in the "lower" end of Lower Fort Street, such as the Harbour View Hotel. The NSWSHR places the hotel in Millers Point, yet the PSMA dataset places the hotel in Dawes Point. The question is, what should we do? Should we use the NSWSHR information or the PSMA dataset? The boundaries of Millers Point are located here; Barangaroo is here; Dawes Point is here; and the City of Sydney is here. I downloaded PSMA dataset of Australian Government official boundaries from here. Thoughts? Rangasyd (talk) 11:05, 14 January 2019 (UTC)
 * Good find. I'd absolutely use the gazetted boundaries over the NSWSHR labels. I corrected quite a few in country areas where the SHR had listed them, for example, in the nearest town, and a coordinates search made it obvious that they were absolutely not in the place where the SHR had said they were. I haven't spent much time in this part of Sydney but I gathered from conflicting search results that similar issues were emerging here. Would be great to see these straightened out and listed as being where they actually are. The Drover&#39;s Wife (talk) 11:17, 14 January 2019 (UTC)

Wikidata
Is there any interest / has any thought been given to the representation of Oz historic places on wikidata - e.g. how to store the heritage register reference ID - as new properties, or as catalog codes pointing to the pertinent national / state register? --Tagishsimon (talk) 00:22, 18 December 2018 (UTC)


 * I think only Kerry Raymond here has any skills with Wikidata - I wouldn't have a clue, sorry. The Drover&#39;s Wife (talk) 00:47, 18 December 2018 (UTC)


 * Thanks. With some digging, I've probably answered part of my own question, at least for NSW. Wikidata has NSW Heritage database ID (P3449), which takes the URL number rather han the listing number - so Darling Harbour Carousel (Q59829948) takes 5053339 rather than 1620 ... I've not yet discovered if there's a pattern for storing that number. --Tagishsimon (talk) 01:01, 18 December 2018 (UTC)
 * In some states (NSW, VIC, WA) there is a register ID and a database ID, and the state register is a subset of the whole database. --Canley (talk) 01:19, 18 December 2018 (UTC)


 * Yes, there has been quite a bit of work done in this area for Wiki Loves Monuments 2018, but still a way to go. Kerry Raymond structured all the registers so they could be imported into (what was at the time) a WLM database, but found out pretty late (at Wikimania) they were using Wikidata instead, so work began on that. I believe all the properties (heritage designations and register IDs) are in place except for the NT register (see wikidata:User:Yarl/Australia for a summary). By the time the properties were set up and Mix-and-Matches and other reconciliation done on the register catalogues, there wasn't enough time to import the rest for the campaign, however I intend to try and do that before the end of the year. I set up the Wikidata-generated lists on Commons here, with links to the state register items where possible. Here is a table of the percentage of items already done:
 * {| class="wikitable"

! State !! Wikidata label !! Wikidata ID !! Number of listed items !! Listings on state register !! Completeness
 * QLD || listed on the Queensland Heritage Register || Q20680290 || 1,753  ||  1,752  || 100.1%
 * NSW || Heritage Act - State Heritage Register || Q28152854 || 749  ||  1,689  || 44.3%
 * WA || State Registered Place || Q56052054 || 687  ||  2,019  || 34.0%
 * VIC || listed on the Victorian Heritage Register || Q28147634 || 753  ||  2,332  || 32.3%
 * SA || listed on the South Australian Heritage Register || Q30166806 || 103  ||  2,297  || 4.5%
 * TAS || listed on the Tasmanian Heritage Register || Q56046814 || 13  ||  6,521  || 0.2%
 * NT || n/a || ||  -    ||  181  || 0.0%
 * }
 * --Canley (talk) 01:16, 18 December 2018 (UTC)
 * SA || listed on the South Australian Heritage Register || Q30166806 || 103  ||  2,297  || 4.5%
 * TAS || listed on the Tasmanian Heritage Register || Q56046814 || 13  ||  6,521  || 0.2%
 * NT || n/a || ||  -    ||  181  || 0.0%
 * }
 * --Canley (talk) 01:16, 18 December 2018 (UTC)
 * }
 * --Canley (talk) 01:16, 18 December 2018 (UTC)


 * Thanks Canley. I saw your fingerprints on one of the newer NSW records & grokked that it was all in good hands. I'm still interested in the 5053339 rather than 1620 NSW ID question - is there a facility/desire for the latter to be stored? It could, I guess, use the catalog_code property. Not greatly important, although I suspect 5053339 is tied to the friable database NSW is currently using, whereas 1620 is the formal & hence persistent ID; clearly we need 5053339 in wikidata to link back to the database, the question is, do we also need or desire the 1620 ID. (My main interest was, from watching Pages not connected to items I see a steady stream of The Drover's Wife's work coming through, and was wondering about how the currently 180 articles without wikidata items should be added. I think that I more or less know now. --Tagishsimon (talk) 01:41, 18 December 2018 (UTC)
 * Yes, both should be properties in Wikidata, I'll submit some proposals to add this and the NT ones soon. This is the case in Victoria where there is a VHR ID and a VHD ID property, but the VHR ID is not linked directly to the website database (it passes the VHR ID as a search parameter). For the new NSWHR items added by The Drover's Wife, I'll check the Mix'n'Match and update these with Wikidata items tonight. --Canley (talk) 02:00, 18 December 2018 (UTC)
 * Experimentally only, https://www.wikidata.org/wiki/Q1529051#P528 --Tagishsimon (talk) 02:09, 18 December 2018 (UTC)
 * this also needs some work - no P31 values for ~550 items. I might chip away at that. --Tagishsimon (talk) 02:30, 18 December 2018 (UTC)
 * That would be great, the "instance of" property is the trickiest part of importing this stuff. Can certainly do with human assessment, maybe the article categories could be some help. --Canley (talk) 02:37, 18 December 2018 (UTC).
 * It's going to be a nightmare. Brush Farm, for instance, seems to have had a gazillion different uses over the years ;) --Tagishsimon (talk) 02:44, 18 December 2018 (UTC)
 * , could the item being re-scoped geographically as a 'instance of: [location or place]' and the vagaries of the title and usage be assigned as 'statements'? cygnis insignis 03:20, 18 December 2018 (UTC)
 * Yeah, I sometimes do those as instance of historic place or historic building rather than listing all the uses. --Canley (talk) 03:49, 18 December 2018 (UTC)


 * (EC) Something like geographic location (Q2221906) is an option as a general purpose P31 (although Canley's suggestions are better). We have use (P366) which will be helpful, for instance, for the listed bank building which can have a P31 of Bank Building and P366s of Bank and Restaurant (since that's what it is now). For now - before thinking about P366 - I've thrown a small handful of P31s at the Farm item; they can perhaps be changed later. FWIW, here are the P31 types for all NSW items that have a P31 ... standarising P31s can also be added to the todo list. Fortunately there is no deadline. --Tagishsimon (talk) 03:56, 18 December 2018 (UTC)


 * That looks better. All of the uses now in P366. https://www.wikidata.org/wiki/Q56277076 --Tagishsimon (talk) 03:59, 18 December 2018 (UTC)


 * Just to note, I've created items for all of the articles without items; but they're all missing their NSW heritage ID right now - they can be found on this report. And I might as well list again the items with no P31 report here too.--Tagishsimon (talk) 20:55, 18 December 2018 (UTC)
 * Thanks for doing that, I'll work on the IDs and instance statements today. --Canley (talk) 22:35, 18 December 2018 (UTC)
 * I'm plodding along adding coordinates to NSW Heritage articles as well, among my other lists of things I'm interested in doing. Not as many, ~70. Link to the query is here Nat965 (talk) 20:42, 23 December 2018 (UTC)


 * Coords are now sorted, bar two items for a collection of locomotives and an illuminated golf-shop sign. But, 520 items with no P31 to tackle next. --Tagishsimon (talk) 00:09, 24 December 2018 (UTC)


 * The golf shop sign is in the Powerhouse Museum collection if you wanted to use those co-ordinates (I am very bad at these things). I couldn't for the life of me work out where the heck the locomotives actually were right now because they've been moved around a bit and it's not terribly well documented which particular train went where. The Drover&#39;s Wife (talk) 00:35, 24 December 2018 (UTC)


 * Thanks TDW. Now its wikidata item has 2 coords, one for 216 Elizabeth Street and t'other for the museum, each with appropriate start/end dates. --Tagishsimon (talk) 01:09, 24 December 2018 (UTC)

Perhaps worth noting that mix'n'match seems to think there are still 97 NSW entries on the heritage list without wikidata items, which points to an absence of articles. I've crosschecked the NSW Register category against wikidata, added items for all articles in it, and added NSW Heritage IDs for all items. Example: Norma Parker Correctional Centre. --Tagishsimon (talk) 01:58, 28 December 2018 (UTC)
 * Oops. mix'n'match link --Tagishsimon (talk) 02:01, 28 December 2018 (UTC)
 * Maybe I am missing something but wouldn't "there are still 97 NSW entries on the heritage list without wikidata items" suggest a problem in Wikidata and not with the articles? Kerry (talk) 08:45, 29 December 2018 (UTC)
 * I think you are missing something, yes, Kerry. I understood from discussion above that this project considered that the NSW Heritage list was complete. That might be my mistake, confusing Sydney for NSW - this diff. I did say in the post to which you are reacting "I've crosschecked the NSW Register category against wikidata, added items for all articles in it, and added NSW Heritage IDs for all items", which was an attempt to forestall a knee-jerk "it must be wikidata's problem". I see I have failed in that. You say in your edit summary "Wikidata has a problem, must be Wikipedia's fault (getting tired of hearing this)". Wikidata does not have a problem, as any reasonable parsing of my comment would indicate. Wikidata does not have a problem because of the hours of recent effort which have gone in to making sure that it keeps up-to-date with en.wikipedia. That being the case, I'm not very sure what to make of your comment, beyond observing that it is unfriendly and unhelpful. --Tagishsimon (talk) 08:56, 29 December 2018 (UTC)
 * In my past experience, the problem of alignment between Wikipedia and Wikidata lie in false assumptions often embedded in Wikidata constraints, that is, they are not modelled correctly. Often they are modelled on the "normal case" and don't realise that there are exceptions. This occurs because nobody ever tests the assumptions with those who might have some topic knowledge. First, know your data! (a principle largely ignored in Wikidata). You can there are 97 NSW entries on the heritage list without wikidata items? Where does 97 come from? Let us unpick this thing. How many of what do we have where? Kerry (talk) 09:29, 29 December 2018 (UTC)
 * I'm afraid I've rather lost any enthusiasm for the NSW project, Kerry. Good luck to you. --Tagishsimon (talk) 09:32, 29 December 2018 (UTC)


 * It is not surprising that there are a number of SHR entries without articles, as a number of entries were merged into shared articles and others were added under different names. For the first five examples: "Joadja kerosene oil shale mining and refining site" is at Joadja, New South Wales as they're the same place, "Glenrock early coalmining sites" I think was never transferred because the SHR entry was useless and we hadn't got into using other CC-BY sources yet, "Norma Parker Correctional Centre" became became Parramatta Female Factory and Institutions Precinct (please nobody redirect it because the prison's still separately notable, the buildings just had a much more complicated history), "Semi-detached Cottages" was merged into Howick Street houses, Bathurst and "Sydney Terminal and Central Railway Stations Group" is at Central railway station, Sydney. I'm not sure what the easiest way of dealing with this is because I'm horrible at Wikidata: we've done Excel spreadsheets explaining what went where for nearly all the sites (we didn't use it for central Sydney), or else I'm happy to just manually explain it. The Drover&#39;s Wife (talk) 09:23, 29 December 2018 (UTC)
 * Thanks for this reply, it's very helpful. All the NSW register items are now reconciled, so the NSW State Heritage Register is now 100% complete in Wikidata—many were not matched before because the names were substantially different or were generic due to the Heritage Office's naming scheme ("House", "Warehouse (former)" and so on). I have created new Wikidata items (not many) where I could not find a corresponding article after considerable effort, or in some cases separate register items which are combined in a Wikipedia article—for example, some of the individual holdings in the Penrith Museum of Fire, or the addresses in the Long's Lane Precinct (these should be linked to the combined topic using the "part of" or "collection" properties). So now the Wikidata Mix'n'Match is complete, would you be able to have a look at this query which lists all the Wikidata items with no corresponding article? There are 54 items, just let me know if any stand out that you think should be linked to an article. --Canley (talk) 12:49, 30 December 2018 (UTC)
 * This is getting a bit complex, so:
 * Some of these I can help with easily: "Watch House" is at Erskine Street Police Station. "Civic Railway Workshops" is at Honeysuckle Point Railway Workshops. "First Government House Site" is at First Government House, Sydney. "APA Building" is at APA Building, Sydney. "Junee railway station moveable relics" went into Junee railway station. "Metropolitan Hotel" is at Metropolitan Hotel, Sydney. "Rhodes railway station waiting shed" is at Rhodes railway station, Sydney. "Samson's Cottage (wall remains)" is at Samson's Cottage wall remains. "Norma Parker Correctional Centre" I noted above. "Sergeant Majors Row (terrace)" is at Sergeant Major's Row, "Callan Park House - Rozelle Hospital" at Callan Park Hospital for the Insane. "Willandra Lakes Region" and "Penrhyn House" are exactly what it says on the box, so not sure why they're showing up there.
 * I am very confused about what's going on in Nowra - I did the Nowra listings (of which there were only two in my list), but there are a ton of listings here which are the first I've heard of them. Something odd has happened here.
 * There are some others that I've never heard before and have no explanation for: "Waratah Park", "Yobarnie Keyline Farm", "Riverview Cottage", "Parramatta Public School", "Illawarra Museum", "Earlwood Aboriginal art site", "Bradfield Carrier", "Air Defence Headquarters Sydney" (the last of which we have two articles for, neither mentioning the SHR). Also very confused about what's happened with these.
 * A couple of the Sydney ones still aren't done yet (Watch House Terrace, anything in Bulletin Place).
 * Lots of this list is in the Penrith Museum of Fire and should link to that article. All the Long's Lane Precinct articles are individually here as well.
 * "Tanilba House" never got written because AussieLegend was throwing tantrums about work in his part of the world and I couldn't be bothered dealing with it. I don't think "Premiers and Railway Commissioners Rail Car Collection" is still in one place anymore and may not have got created because I couldn't nail down where the heck everything had gone to. "Lord Howe Island Group" looks like it was never transferred, probably due to not fitting into Lord Howe Island, but was merely mentioned there - Wikidata should point there anyway.
 * Hope this helps a bit! The Drover&#39;s Wife (talk) 21:21, 30 December 2018 (UTC)
 * Thanks for the help, that's awesome! Some responses:
 * I have merged any duplicated items you noted above. "Penrhyn House" was incorrectly linked to a house in Wales!
 * Don't worry about the Nowra ones, I'll check more thoroughly but I don't think they are on the SHR, just on an LEP so they are in the heritage database but not in scope for the state register.
 * Some of those names are very different to the registered name and reflect the current usage: "Parramatta Public School" = "King's College (former)"; "Illawarra Museum" = "Original Wollongong Telegraph and Post Office". I saw one ("Riverview Cottage") that had been delisted from the register in 2016.
 * I mentioned the Penrith Museum of Fire and Long's Lane Precinct items (also Parramatta Female Factory and Institutions Precinct). I'll keep these as individual Wikidata items for now, as they can be linked to the collective item and article.
 * I created an item for "Premiers and Railway Commissioners Rail Car Collection"—as you said, it seems to be all over the place. The coordinates in the register entry are in Eveleigh, but the cars seem to be distributed between Broadmeadow (Newcastle) and Thirmere (Wollondilly).
 * Thanks again and happy new year! --Canley (talk) 23:07, 30 December 2018 (UTC)
 * Ah, that clears up a bunch - but I'm still baffled about that random group that I've never heard of (apart from Riverview if it's been deleted). "Parramatta Public School" seems like it must refer to Old King's School, Parramatta even though the Parramatta school is only temporarily operating out of there during construction works on their original site. We don't have an article for "Original Wollongong Telegraph and Post Office" and I still dunno about the others. The Drover&#39;s Wife (talk) 00:40, 31 December 2018 (UTC)
 * Ah yes, the King's School one was because the NSWHR has two items for what appears to be the same site ("Marsden Rehabilitation Centre Group" and "Kings School Group (former)") at different addresses (O'Connell Street and Marist Place). I've left "Parramatta Public School" as a separate school item as it will likely move around. Here's a new query which excludes the items part of a collection or collective group. The items not previously mentioned, which might be included the random group, are "Cootamundra No. 3 AIFD" (WW2 and later civilian fuel depot), "Newcastle Government House" (seems to be a parade ground and James Fletcher Hospital now) and "Rose Bay seawall"—do they ring a bell? --Canley (talk) 01:01, 31 December 2018 (UTC)
 * Also baffled about those three - and I did Cootamundra and Newcastle so they definitely weren't on my list. The Drover&#39;s Wife (talk) 01:08, 31 December 2018 (UTC)

One of the many things on my to-do list before I go on holidays is to catch up on merging the individual LGA spreadsheets that TDW and RangaSyd have been updating during the NSW SHR rollout into the master spreadsheet. That should then be compared against the Wikidata list to reconcile them and then we can explore the mysteries. I just don’t know if time will permit though (a number of real world matters are consuming my time). Kerry (talk) 07:25, 31 December 2018 (UTC)
 * I can do the Wikidata reconciliation, but let me know if you need any help with the spreadsheet merging as well. --Canley (talk) 11:44, 31 December 2018 (UTC)
 * I'm just picking this up now. Sorry, I was not following discussion earlier. I've just run the above query and this is the result, together with my comments/suggestions. We might have to wait until Kerry returns to address those that did not auto-generate. Comments from me are below: Rangasyd (talk) 12:16, 14 January 2019 (UTC)
 * {| class="wikitable"

! Wikidata ID !! Wikidata label !! NSWSHR ID !! Status / comments
 * Q60233495 || 10–14 Bulletin Place, Sydney || 5045674 ||rowspan=3|Incomplete. An article named Bulletin Place warehouses to be created to cover all three items
 * Q60233535 || 16–18 Bulletin Place, Sydney || 5045084
 * Q60231981 || Bulletin Place Restaurant || 5045669
 * Q60232199 || Cootamundra No. 3 AIFD || 5062423 || Not listed in the auto-generator from Kerry
 * Q60232377 || Earlwood Aboriginal art site || 5060975 || Recommend redirect to Baiame Cave
 * Q60228235 || Illawarra Museum || 5062291 || Not listed in the auto-generator from Kerry
 * Q46775862 || Lake Parramatta Dam ||	5056348 || Recommend redirect to Lake Parramatta
 * Q1869866 ||	Lord Howe Island Group || 5001478 || As per The Drover&#39;s Wife above, recommend redirect to Lord Howe Island
 * Q60233196 || Newcastle Government House || 5060998 || Not listed in the auto-generator from Kerry
 * Q60220255 || Riverview Cottage || 2680130 || This is NOT a NSW SHR item. It's a NSW HD item. NSW HD is outside the scope of this project initiated by Kerry. I recommend delete the Wikidata item.
 * Q60218874 || Rose Bay seawall || 5055617 || Not listed in the auto-generator from Kerry
 * Q60229346 || Tanilba House || 5045714 || See The Drover&#39;s Wife comments above. Article to be created was titled Tanilba House and The Temple
 * Q60227907 || Waratah Park || 5061163 || Not listed in the auto-generator from Kerry
 * Q60228290 || Watch House Terrace ||	5045343 || As per The Drover&#39;s Wife above, recommend redirect to Erskine Street Police Station
 * Q60227808 || Yobarnie Keyline Farm || 5061775 || Not listed in the auto-generator from Kerry
 * }
 * Q60233196 || Newcastle Government House || 5060998 || Not listed in the auto-generator from Kerry
 * Q60220255 || Riverview Cottage || 2680130 || This is NOT a NSW SHR item. It's a NSW HD item. NSW HD is outside the scope of this project initiated by Kerry. I recommend delete the Wikidata item.
 * Q60218874 || Rose Bay seawall || 5055617 || Not listed in the auto-generator from Kerry
 * Q60229346 || Tanilba House || 5045714 || See The Drover&#39;s Wife comments above. Article to be created was titled Tanilba House and The Temple
 * Q60227907 || Waratah Park || 5061163 || Not listed in the auto-generator from Kerry
 * Q60228290 || Watch House Terrace ||	5045343 || As per The Drover&#39;s Wife above, recommend redirect to Erskine Street Police Station
 * Q60227808 || Yobarnie Keyline Farm || 5061775 || Not listed in the auto-generator from Kerry
 * }
 * Q60227907 || Waratah Park || 5061163 || Not listed in the auto-generator from Kerry
 * Q60228290 || Watch House Terrace ||	5045343 || As per The Drover&#39;s Wife above, recommend redirect to Erskine Street Police Station
 * Q60227808 || Yobarnie Keyline Farm || 5061775 || Not listed in the auto-generator from Kerry
 * }
 * Q60227808 || Yobarnie Keyline Farm || 5061775 || Not listed in the auto-generator from Kerry
 * }
 * }

NSW State Heritage Register - ready to roll?
I think we are ready to start to roll out the articles. I think the generator as good as it can get until I can get some feedback on the rollout of real articles. So if you are ready, willing and able to receive some drafts, let me know and we can discuss how best to give them to you. Kerry (talk) 01:04, 28 May 2018 (UTC)
 * I've now done Yass Valley Shire, City of Albury, Armidale Dumaresq Shire, Municipality of Ashfield and working my way through the more-complicated City of Auburn if anyone wants to take a look - always useful to have some further eyes on these. The Drover&#39;s Wife (talk) 10:41, 2 June 2018 (UTC)
 * Now that we've all got all the files, so Rangasyd and I don't trip over each other - I'm currently doing Bankstown and Bathurst (and the one-article-each Ballina and Balranald). The Drover&#39;s Wife (talk) 01:07, 7 June 2018 (UTC)
 * I have added the table below as an easier way to track who is working on what so you don't get under one another's feet. I believe the table reflects the current situation wrt to each of you, but obviously edit it if that's not correct. If there;s a better place to put this table, please negotiate on where. Total is the number of heritage entries for that LGA, which is then broken down on how many are likely to be pre-existing articles and how many are likely to be new articles to give you some sense of the situation before you pick an LGA. But I only did a quick scan for the pre-existing articles so this breakdown may not be 100% accurate, but the total should be accurate). Kerry (talk) 02:19, 7 June 2018 (UTC)
 * Thanks for adding this - very helpful. The Drover&#39;s Wife (talk) 02:33, 7 June 2018 (UTC)
 * I don't seem to be able to edit the Google Docs CSV files after all - I can open them in Google Sheets, but it doesn't save them in the same place. The Drover&#39;s Wife (talk) 03:49, 7 June 2018 (UTC)
 * I take that back - it seems to be creating duplicate sheets in the same space that I can edit? Can you see these Kerry? The Drover&#39;s Wife (talk) 04:34, 7 June 2018 (UTC)
 * Yes, I can see both my original and your modified CSV. While this isn't what I was expecting Google Drive to do, it's quite acceptable (even desirable) for our purposes. Kerry (talk) 16:22, 7 June 2018 (UTC)
 * one listing had a wiki note placed on a word that was wrong context, one more thing to be looking for, a persons surname was linked to a suburb named after him. Minor Dave Rave (talk) 05:20, 18 August 2018 (UTC)
 * the sub paragraph headings has a typo issue, one confuses up the is ihas and it's been promulgated to nearly every entry. Dave Rave (talk) 07:12, 8 October 2018 (UTC)
 * - now that someone else mentions it, if it's not too late, it'd be great to get rid of that "ihas" typo bug, which has been driving me mad all along. Not sure I understand the issue about the lead heading paragraph, Dave Rave? The Drover&#39;s Wife (talk) 07:20, 8 October 2018 (UTC)
 * Fixing it now. Someone should have told me sooner. I will advise when the fixed drafts are uploaded (over 800 drafts affected) 07:44, 8 October 2018 (UTC)
 * Done. Your NSWSHR drafts folder should now be updated. What takes the time is uploading them all to Google Drive! Kerry (talk) 09:07, 8 October 2018 (UTC)

- as per your recent edits - as far as I can tell, they're divided by old LGA because the NSWSHR hasn't been updated with the new ones. The generated text, however, is showing the current, correct LGAs - so not sure there's a need to note all the individual mergers. The Drover&#39;s Wife (talk) 21:32, 8 June 2018 (UTC)

Listings with minimal information and aggregation of multiple listings
I've run into the problem we discussed a lot earlier with regard to Millers Point and the number of "House" listings with minimal information in Bathurst: I merged two of them together as Bentinck Street houses, Bathurst. Any thoughts are welcome.
 * I also had this problem with another bigger group of houses - Howick Street houses, Bathurst - but that got so complicated at five houses, four heritage listings and only two with any information that I raided the local heritage register to fill in the blanks for the other houses (doing manual attribution in the absence of the generator for it yet).The Drover&#39;s Wife (talk) 12:28, 9 June 2018 (UTC)
 * Another issue - I'm having a bit of an issue with the Bathurst articles because the SHR uses a cut-and-paste history of the founding of Bathurst in each and every article, even when it's really irrelevant to the subject at hand. It looks a bit odd when the text has paragraphs about the founding of Bathurst and then a huge time gap until it actually gets to the subject of the article - especially on the shorter articles. I'm trying to use a bit of judgment about when to take them out - e.g. it makes sense to have the blurb there for the pastoral properties or early colonial buildings that actually go back to that era, but the detailed founding of Bathurst really has not much to do with Ben Chifley's House or the Bentinck Street Elm Trees. The Drover&#39;s Wife (talk) 05:39, 10 June 2018 (UTC)
 * Ummm, I'm originally from Bathurst. I can apply some local knowledge if you want me to review. And please let me know if you want some photos taken. I can do so when next back in town. Although, the elm tress would look spartan at this time of year. Rangasyd (talk) 11:48, 12 June 2018 (UTC)
 * Both the local knowledge and the photos would be fabulous. We already had photos for a handful, but there's a lot that still need them - Bathurst is a particularly big group of listings for an LGA. The Drover&#39;s Wife (talk) 13:29, 12 June 2018 (UTC)
 * Ok. I will get to it next month. Rangasyd (talk) 08:03, 15 June 2018 (UTC)

- on the issue of listings with minimal information, Kerry suggested by email just before she left that we note any articles that need to be topped up with the local/other state heritage database content in the "issues" column, so it's easier to come back and top up some of the cruddier ones. The SHD often has far better content for these minimalist articles that lack a history or a description - see this entry for Grenfell railway station as compared to the SHR listing. The Drover&#39;s Wife (talk) 22:51, 16 June 2018 (UTC)
 * Thanks. We just need to be careful to understand that this information is not why it was listed on the SHR. Looking specifically at the entry for Grenfell railway station, the entry has been created by virtue of s.170 of the Heritage Act whereby any property owned by a NSW Government agency may automatically list a property on the NSW State agency heritage register. This register is different to the New South Wales Heritage Register. It's complicated, but there are basically three levels of the NSWSHR;
 * Those that are approved by the NSW Government under the Heritage Act as being of State (and/or national/international) significance;
 * Those that are listed by virtue of s.170 by a NSW Government agency, that have significance to the agency, but their significance to the State has not been approved (yet). With these, the railways properties have been the most active; and
 * Those that are listed by various local government authorities, that have no state significance (yet).
 * In summary, I'm comfortable using the content from those covered by s.170, but less comfortable when using local government information, as it varies significantly from LGA to LGA. I've made a note to Kerry's CSV file for articles to review later (re issues column). Keep up the great work. Cheers. Finally, would you please PM me with a list of Bathurst properties that you need photos for and I will action them when visiting in the next few weeks. Rangasyd (talk) 16:44, 17 June 2018 (UTC)
 * We derive the notability of these articles from the Heritage Register, and we use their specific statements as to significance, but I'm not sure the rest of the article (history/description/etc) needs to reflect the heritage register where it isn't great (and already doesn't where we have existing articles). We've already got this additional great CC-BY content outside the entries themselves - may as well use it. The Drover&#39;s Wife (talk) 22:02, 17 June 2018 (UTC)
 * As for Bathurst (may as well do here rather than figure out PMs again). The SHR ones without articles are Miss Traill's House, Old Government Cottages Group, Bathurst, Bathurst Showground, Bathurst Rail Bridge over Macquarie River, Royal Hotel, Bathurst, Howick Street houses, Bathurst, Merembra Homestead, Bathurst Street Lamps , Bentinck Street houses, Bathurst, Bentinck Street Elm Trees, The Grange and Macquarie Plains Cemetery. Some of these are more important than others, obviously, but any you could grab while you're down there would be brilliant. The Drover&#39;s Wife (talk) 22:02, 17 June 2018 (UTC)
 * I will action the above as they're done. There are lots of street lamp images ... just needed local knowledge on what to look for :-). Where articles do not have images, I've been tagging as Photo requested (in=New South Wales or in=Sydney). I am also using The to start an article, were required; and are using |format=dms for the coordinates. Rangasyd (talk) 16:47, 19 June 2018 (UTC)

Railway stations
Which infobox should we use for historic sites that are also railway stations? Infobox station and incorporate the designation, or Infobox historic site, where there is no capacity to incorporate the railway station information, requiring the need to establish a seperate infobox for the station information. An example of the former is at Albury railway station and examples of the later is at Dunedoo railway station and Dumaresq railway station. My preference is for the former. Rangasyd (talk) 08:03, 15 June 2018 (UTC)


 * I've tended to keep the existing station infoboxes for the current stations and use the historic site ones for the former stations that I'm adding for the first time. I don't have a strong opinion either way - comparing the two at Dumaresq, it seems that no information is being lost by using the station infobox rather than the historic site one and it contains other useful information left out of the historic site one. On the other hand, I'm not a big fan of fiddling with infoboxes and having to replace them - haven't really had to use the station one before. The Drover&#39;s Wife (talk) 08:16, 15 June 2018 (UTC)
 * Let's just use Infobox historic site for new articles; and at a later stage we can go back and merge with Infobox station if we think it's appropriate. For any historic site with an existing article and a Infobox station, I recommend that we use the existing infobox, namely Infobox station. Rangasyd (talk) 13:04, 15 June 2018 (UTC)
 * That sounds great to me. I'm about to do two more when I work through Berrigan Shire today. The Drover&#39;s Wife (talk) 21:19, 15 June 2018 (UTC)
 * it’s ok to have 2 info boxes. If it’s an active railway station, put that one first. Or, name the heritage article “Smallville railway station buildings” and cross link the two articles. I do that a lot when merging an existing article that is focussed on the function of the buildings rather than the buildings, eg Brisbane Boys' College Buildings. Kerry (talk) 13:47, 17 June 2018 (UTC)
 * I've always felt the double infobox looked very ugly and was something to be avoided if possible, and in this case the station infobox allows us to include basically everything of substance with no loss of information. The Drover&#39;s Wife (talk) 00:03, 19 June 2018 (UTC)
 * double boxes are to be avoided where possible, but getting the "module =" working is not very consistent and just a lot more work, too. ANd I've yet to find the use of an infobox other than to display the details neatly in a box. Does the info also get collated elsewhere ? Dave Rave (talk) 05:18, 18 August 2018 (UTC)

Owners
I've just twigged that there might be an issue with the owners not being updated - considering that they're frequently private individuals or companies, I'm not sure we should be using that data. I'm just creating Exeter Farm, which the SHR entry acknowledges was sold to a private buyer in 2013, and it still lists the Office of Environment and Heritage as the owner. If the ownership more broadly is potentially that out of date (which I'm assuming it is, judging from some other articles I've created), this could be a problem. Not so much where the owner is clearly government/NGO/something obvious, but at least for things in private ownership I feel like we might want to remove it. The Drover&#39;s Wife (talk) 23:02, 16 June 2018 (UTC)
 * Agree. I was reflecting on that info a few days ago. Let's omit. Rangasyd (talk) 12:15, 17 June 2018 (UTC)
 * Or leave it be and hope someone local corrects it. You do have a source to justify including what is in the NSWSHR. If you definitely know someone else owns, then obviously fix it. But if you are just speculating it may no longer be correct, then I would say leave it as the NSWSHR has it. Kerry (talk) 13:52, 17 June 2018 (UTC)
 * I also think that the ownership description is clumsy; e.g. "The property is owned by Rail Infrastructure Corporation (State Government)". I substitute with "The property is owned by the RailCorp, an agency of the Government of New South Wales." (RIC no longer exists). Rangasyd (talk) 16:47, 19 June 2018 (UTC)
 * I also like this a lot more than the default and really should use it myself. The Drover&#39;s Wife (talk) 22:44, 19 June 2018 (UTC)

Listing in LGAs as well as suburbs; and use of Sydney map
I've been listing the heritage sites in both the LGAs and in the suburbs. Yet I've notice that you both haven't. I'm not too sure why I started, but I think it is of value. Thoughts? Also, for those heritage sites located in greater metropolitan Sydney, I've been using the "Australia Sydney" locmap instead of the "New South Wales" locmap; as I feel it's more descriptive of the heritage item and it's relative placement. Cheers. Rangasyd (talk) 13:42, 11 July 2018 (UTC)
 * I don't have a strong opinion about the LGAs - not sure it's relevant but happy to do it if people think so. The Sydney locmap makes sense - how do I do that? The Drover&#39;s Wife (talk) 23:08, 11 July 2018 (UTC)
 * I’d been thinking of doing LGA lists for the QHR sites, someone (I think Mattinbgn) created a few manually, but I just hadn’t got around to doing the others. It’s something that could be generated (for Qld and NSW) if we could agree on the format. So if you want to think about your preferred format (including how to order them), I can do the generation when I get home late July. I don’t think it is should be hard to generate the content of each entry, but putting them into the desired oreder might be more of an issue. IIRC Mattinbgn used tables which gives us the potential to sort on multiple columns eg name, suburb, but trickier (impossible) to sort by street name. Or we can use a list format as we do for the suburbs, in which case I would be inclined to have an alphabetical list of suburbs with sub-lists by address (as in the suburb articles. And you may have other ideas ... you may also have to prioritise LGA listings bs mini-generators :-) As for maps, whatever you think best for the reader. I think we can set it up to allow the reader to switch between maps if you like and offer both. Someone (not me) set up the Qld ones to be shown optionally with Qld or Australian maps. Kerry (talk)
 * I prefer the lists over the tables I think (as for the suburbs). Generating it when you get back sounds like a good plan. The Drover&#39;s Wife (talk) 22:09, 12 July 2018 (UTC)
 * Where Kerry's autogenerator populates ; substitute with . Check to make sure that the pin falls within the map coverage area (Hawkesbury/The Hills, Macarthur, Sutherland/Illawarra, lower Blue Mountains regions are likely impacted). Rangasyd (talk) 09:15, 13 July 2018 (UTC)
 * A few options for you: List of rivers of New South Wales (A–K); Category:Lists of historic places. I quite like the List of New York State Historic Sites. My thoughts are that the layout will need to be laid out in a table and include sortable fields such as (in no order) heritage register id number, "date of establishment", LGA, title, location (suburb/city), coords, image. I also feel that we need to break down Category:New South Wales State Heritage Register. There are 400+ entries thus far and we're only halfway there. For example, in Victoria, there is a category for Category:Heritage listed buildings in Melbourne, which I feel is a little vague (can only list buildings, what about heritage sites in Melbourne that are not buildings (e.g. cemeteries, shipwrecks, etc). The equivalent would be "Category:New South Wales State Heritage Register sites located in Greater Sydney." Rangasyd (talk) 09:15, 13 July 2018 (UTC)

Bridge nomenclature
Acknowledging that the NSW SHR typically (but not always) names bridges in the format XXX (town/locale) Bridge over (the) YYY River, in common practice in Australia we typically say/write as YYY River bridge, XXX (town/locale). The only exceptions to this is where the bridge has an official and commonly used name, e.g. Sydney Harbour Bridge, Anzac Bridge, Denison Bridge, Captain Cook Bridge, New South Wales, etc. There is also over-capitalisation on use of the word bridge. The bridges that I wish to draw to your attention are listed below, with suggested replacement title:
 * Barham Bridge over Murray River→Murray River bridge, Barham (completed)
 * Bridge over Turon River at Wallaby Rocks→Turon River bridge, Wallaby Rocks (subsequent discussion to rename as Wallaby Rocks Bridge)
 * Carrathool Bridge over Murrumbidgee River→Murrumbidgee River bridge, Carrathool (completed)
 * Clarence Town Bridge over Williams River→Williams River bridge, Clarence Town (completed)
 * Coonamit Bridge over Wakool River→Wakool River bridge, Coonamit
 * Cooreei Bridge over Williams River→Williams River bridge, Cooreei
 * Gee Gee Bridge over Wakool River→Wakool River bridge, Gee Gee
 * Greigs Flat Bridge over Yowaka River→Yowaka River bridge, Greigs Flat (completed)
 * Monkerai Bridge over Karuah River→Karuah River bridge, Monkerai (completed)
 * New Buildings Bridge over Towamba River→Towamba River bridge, New Buildings (completed)
 * Rossi Bridge over Wollondilly River→Wollondilly River bridge, Rossi Crossing
 * Swan Hill-Murray River Road Bridge→Murray River road bridge, Swan Hill (completed)
 * Vacy Bridge over Paterson River→Paterson River bridge, Vacy (completed)
 * Wee Jasper Bridge over Goodradigbee River→Goodradigbee River bridge, Wee Jasper (completed)


 * I like this convention a lot and I've been using it to create new bridges since you did the first batch of moves (excepting some like Queanbeyan railway bridges over Queanbeyan and Molonglo Rivers where I wasn't sure what to call it). However, in some of these the first noun seems to be a name and not a place - for example, there is no town/locality called "Cooreei" (it's in Dungog), none called "Coonamit" (on the border of Mallan and Dilpurra), none called "Wallaby Rocks" (on the border of Sofala and Crudine), none called "Gee Gee" (on the border of Wakool and Cunninyeuk) and none called "Rossi Crossing" (Goulburn) - not sure of the origin of the name in each case. The Drover&#39;s Wife (talk) 21:48, 28 July 2018 (UTC)
 * add GNB to your list of tools to use in NSW, we use localities and feature names too Dave Rave (talk) 05:27, 18 August 2018 (UTC)
 * Thanks Dave Rave. Rangasyd (talk) 04:14, 20 January 2019 (UTC)
 * there are six crossings of the Wollondilly at Goulburn, and it's a long river. I could get used to the new style, but I don't like it. Dave Rave (talk) 17:35, 20 January 2019 (UTC)
 * I don't have any problem using other common names if they exist in cases like that - this is most useful where there is a bridge over a river at a place but no particular name and it gets all of those articles on the same page without making stuff up/using eight different styles. For example, at the Turon River bridge at Wallaby Rocks (which is one of the handful that didn't quite fit in the section above), I suggested going back to your original name of Wallaby Rocks Bridge (where it would have stayed if I'd found the existing article before creating the SHR one). The Drover&#39;s Wife (talk) 21:10, 20 January 2019 (UTC)
 * Regarding the Bridge over Turon River at Wallaby Rocks, at the article's talk page there is a proposal to rename the article. I encourage editors to please direct all discussion on the renaming of this article to the talk page and not here. Many thanks. Rangasyd (talk) 14:15, 21 February 2019 (UTC)

as for the rest of these - why not use the same approach as we just did at Wallaby Rocks? Rather than saying they're in a nonexistent locality (or in some cases nonexistent place at all), there are generally sources for the remaining ones at (for example) Rossi Bridge. The Drover&#39;s Wife (talk) 22:49, 25 February 2019 (UTC)

Ref order
I'm not overly worried by the refs showing up as [24][1], instead of [1][24], but others would know where to point you to a wiki article saying the first one should be first. Doing the manual tidy can be long and tedious and one thing less to do would be nice. Dave Rave (talk) 09:30, 12 September 2018 (UTC)
 * Hi there. It looks untidy to me too. Kerry, can you please fix there auto generator for future projects so that the smaller reference number appears before the larger reference number. I think you default is that, for example, the NSW SHR reference appears at the end of every paragraph. Typically this reference will be [1]. Is it possible that this reference appears before any other reference that will be numbers higher than [1]? Cheers. Rangasyd (talk) 13:00, 5 October 2018 (UTC)
 * The problem which we have inherited from traditional print is that the scope of a citation is not precisely indicated. All that we know is that the citation supports the text before it, but how much text? A sentence, two sentences, a para? Now on Wikipedia we *could* do better by linking the citation to one or more text spans that it supports, which we could display to the interested reader through colour highlighting or suchlike. Imagine if we wrote markup like this

So the [1] supports the whole paragraph, so it must go last. While we do not know with certainty the text span related to the [24], it is most likely the later part of the paragraph, a subset of that supported by [1]. It is only if we know the span of [24] is the also whole para can we reorder as [1][24], otherwise we subtly change the meaning. This is not a question of aesthetics but rather one of semantics, albeit with degraded meaning due to our failure to fully exploit the advantages digital has over paper.

The problem is particularly compounded in a loose collaborative nature of a wiki. In an academic paper, there is a single author or multiple authors who are in close collaboration. The authors all understand the content of the paper and they all know the intended span of the citation. But in the loose stigmergic collaboration of a wiki, when we rewrite an existing chunk of text, we face the problem of where to reposition the existing citations in the rewritten text as we don’t know the original span intended by the person who added the citation. As a result, we get what I call “citation drift” where the citation moves away from the text it supports. Over time, the citation may get deleted as a possibility. If we were forced to make explicit the span of the citation, this would be less of a problem.

Does this make sense? Kerry (talk) 15:14, 5 October 2018 (UTC)
 * BTW, the generator does not know the numbers of the citations, it just names them no different to someone manually writing an article; the numbering [1][2] is done by the wiki-rendering when someone reads the article. Any reordering of text or earlier reuse of a citation produces out-of-order numbering so this is normal Wikipedia behavior and nothing to do with generation, so I would really be surprised if we could have a policy about it for that reason. The sequencing of the numbers is important in the reference list but not in the text. Similarly Itwould be possible to reorder them in the render but I presume they don’t for the semantic reason. Kerry (talk) 15:30, 5 October 2018 (UTC)
 * Yes it does make complete sense and I would definitely support such as a principle. My concern is would editors cope with non trivial tracts of text being "bracketed up" and very often nested a few or more layers deep.  It really would be directly analogous to variable scoping in coding.  While references are presented at the end, perhaps the "cite" should go at the start of the text block(span) template (the same as a variable declaration), making it easier to see what the (following) text is based on?  Aoziwe (talk) 16:53, 5 October 2018 (UTC)
 * I don't think it matters how we would build it. There isn't a serious proposal on the table to introduce citations with spans. I am familiar with the issues of lack of spans because it comes up from time to time on the research mailing list. People who want to (say) build a tool to find uncited facts in an article quickly run into the problem. Given we have plenty of editors who don't provide citations at all, I think citing with spans would be very hard to introduce. It would be *better* but we have to stay within the abilities of the average contributors (especially at a time when we need more of them). It would be far easier to introduce with the Visual Editor where we could just ask the editor to select the blocks of text supported by the citation they just added. It would be much more difficult for anyone using the source editor. Kerry (talk) 05:23, 6 October 2018 (UTC)
 * In fact this issue about ordering of citations is discussed in WP:INTEGRITY and, as I suspected, does not recommend reordering them for fear of breaking the source-text relationship. Kerry (talk) 05:33, 6 October 2018 (UTC)
 * Yep, I do get it .. C'est la vie. Aoziwe (talk) 13:59, 6 October 2018 (UTC)

Blank indigineous section
if it doesn't have a worthwhile aboriginal paragraph, can the lead heading paragraph be left off ? Dave Rave (talk) 07:12, 8 October 2018 (UTC)
 * Can you point me at an example of an article (or an earlier version of an article) that has this problem with the aboriginal paragraph? (I am not quite sure what problem you are seeing) Kerry (talk) 05:14, 9 October 2018 (UTC)
 * the bit about aboriginal history, sometimes it has been left blank after the heading. Dave Rave (talk) 03:30, 9 October 2018 (UTC)
 * I looked through my watchlist and opened up everything possible, but I can't find it. But I did see it. Might be a topic not on my watchlist, but I checked contributions too. Dunno
 * I'll work up an example in my sandbox Dave Rave (talk) 05:42, 11 October 2018 (UTC)
 * Found one - Davisville, Wentworth Falls Dave Rave (talk) 22:18, 27 October 2018 (UTC)
 * Ah, that's not generated. It is something that User:Rangasyd added. Kerry (talk) 22:48, 27 October 2018 (UTC)
 * Another - Guildford West pipehead and water supply canal added from creation Dave Rave (talk) 22:57, 30 October 2018 (UTC)
 * Again, this is not generated. It is not in the generated draft article. User:Rangasyd must have added it before doing the first Save. I presume Rangasyd is adding this empty section in order to encourage people to add to the Indigenous history of this article. Some of the NSW SHR entries do contain Indigenous history but most don't. Kerry (talk) 00:09, 1 November 2018 (UTC)
 * One of the big issues with the NSW SHR is that in the main its written from a white fella perspective of history and there is little consideration given to the history of the place prior to 1788. 11:43, 2 November 2018 (UTC)
 * Just exampling another, so you can search for more, Ashfield Reservoir plus the hyphen on the next heading Dave Rave (talk) 20:28, 17 November 2018 (UTC)

Use of ibid in refs
I know it's a chore, like the number/convert option, but reading through Paddington Reservoir was, for me (who does refs) annoying. (Not that I'm angry, just that I've learnt there is a way to do this and we frown on wp:ibid) Dave Rave (talk) 22:25, 27 October 2018 (UTC)


 * No disagreement here. I usually only fix it if there's a ridiculous number of ibid citations because messing too much with templates breaks my brain a bit and I'm always pleased when someone who's a bit more comfortable with this stuff comes along and fixes the remainder. The Drover&#39;s Wife (talk) 22:50, 27 October 2018 (UTC)


 * The generator knows about ibid and, if it sees them in the heritage register, it will reuse the previous citation as in so most of the ibids are resolved and you don't see them in the wikitext. But the problem comes when the heritage register says "ibid, blah blah", which means "the last citation except this time with blah blah" ("blah blah" is often a page number). The generator always knows what its previous citation was and it knows it needs "blah blah" substituted into that previous citation (which from a Wikipedia perspective is easiest done by creating a new citation with the substituted text) but the generator doesn't know what part of the previous citation it is supposed to be replacing. So it gives the human rolling it out the information they need, by generating a new citation "previous citation text // ibid, blah blah" and expects the human to use their intelligence to work out how to modify the previous citation text with "blah blah". Now in the Queensland Heritage Register, ibid was quite rare. It's a lot more frequent in NSW SHR. Indeed, I just did some quick counts. Ibid appears in just over 250 NSW SHR entries. The generator resolves these successfully in about 200 articles, leaving ibids in about 50 articles for human resolution. Paddington Reservoir is one of those 50. Now the problem with unresolved ibids is that they can cascade. If the text contains "... [some citation] .... [ibid, this] ... [ibid, that] ...", then the generator cannot be entirely sure if the 2nd ibid is dependent on the modification done for the first ibid, so it cascades them by saying "some citation // ibid this // ibid that" so the human has full information to resolve it. This is why you see in Paddington Reservoir this cascading sequence of ibids " " which was of course unnecessary. If the NSW SHR had just said "ibid" each time (as it didn't want to modify the original citation in any way), it all would have resolved itself within the generator, but because they insisted on repeating the page number each time, the generator thinks some modification is needed and is forced to hand the problem to a person to resolve. Most ibids are trivially resolvable by a human. I'll go through and fix up any ibids still in NSW SHR articles, starting with Paddington Reservoir. Kerry (talk) 23:51, 27 October 2018 (UTC)
 * Having said that "most ibids are trivially resolvable by a human", I did of course mean to say "when they were intelligently added in the first place". As I am going through fixing them in the NSW SHR articles, the biggest problem is that the use of ibid is seriously broken in the original heritage register entries. It looks like the SHR entries have been edited by different people over time and in the process have added new citations that separate the ibid from whatever it was supposed to be "same as" (which is of course why we don't want ibid in Wikipedia). It's proving quite challenging to unscramble some of these. Kerry (talk) 19:53, 28 October 2018 (UTC)
 * Mea culpa, mea culpa, mea maximuma culpa. Rangasyd (talk) 13:16, 1 November 2018 (UTC)

Status update on the New South Wales State Heritage Register project
I thought folks might be interested in seeing where things are at with this project. I think progress has been incredibly impressive. Given that I think the first real work happened in March 2018, a year later we now have articles on almost all of the NSW SHR properties. Note that multiple adjacent heritage properties (e.g. rows of terrace houes) sometimes get combined into a single Wikipedia article so there will always be slightly fewer Wikipedia articles than heritage listings. The other good news is that about 2/3rds already have photos on Commons to use. The NSW project been many times faster to get to this point than with the earlier Queensland Heritage Register project. This is predominantly due to having 3 people rather than just one and I must pay tribute to and  for their relentless pace during the roll-out. I don't know if they realised just how much work they were signing up to do, but they never faltered in their enthuasiasm. Of course, I re-purposed a lot of code from the Queensland project and of course I learned from my mistakes on the Queensland project, both of which are time savers. Kerry (talk) 07:22, 21 March 2019 (UTC)

Participants

 * Rangasyd (article roll-out, photography)
 * The Drover's Wife (article roll-out)
 * Kerry Raymond (code monkey, categorisation)

Statistics for en.Wikipedia (as at 23 March 2019)

 * Number of sites listed on the New South Wales State Heritage Project when we started: 1,695
 * Number of existing articles about those topics on English Wikipedia when we started: 383
 * Number of articles in Category:New South Wales State Heritage Register (and sub-cats where appropriate): 1,642
 * Number of new articles: 1,299
 * Number of expanded pre-existing articles : 383
 * Number of articles with an infobox photo: 1,032
 * Number of Commons categories as sub-cats of Category:New South Wales State Heritage Register: 1,018

All statistics are approximate.

Next phase: upload suitably licensed images from the New South Wales State Heritage Database to Wikimedia Commons: estimated 6,000 photos, hopefully including some for the approx 600 articles currently without photos. It is hoped this can be largely automated using the Pattypan tool. More manual tasks will include the creation of the Commons categories as needed, and adding photos to the articles.

Subsequent phase: add Photo requested tags to all the articles still without photos.

A new newsletter directory is out!
A new Newsletter directory has been created to replace the old, out-of-date one. If your WikiProject and its taskforces have newsletters (even inactive ones), or if you know of a missing newsletter (including from sister projects like WikiSpecies), please include it in the directory! The template can be a bit tricky, so if you need help, just post the newsletter on the template's talk page and someone will add it for you.
 * – Sent on behalf of Headbomb. 03:11, 11 April 2019 (UTC)

List of historic homesteads in Australia
Hi all. Sorry if you are on top of this but with all the relatively recent activity on various heritage registers, should there have been more activity on List of historic homesteads in Australia too, or has it slipped through a wikicrack? Regards. Aoziwe (talk) 12:35, 18 February 2019 (UTC)
 * The big questions with the list are:
 * What comprises a homestead? Does it differ from a house? When does a house become a homestead?
 * This list requires manual updating and is not dynamic. How do we sustain maintenace?
 * Cheers. Food for thought. Rangasyd (talk) 14:32, 21 February 2019 (UTC)
 * I do not know, but if the artilce is agreed to be there then it needs to be supported. Maintenance itself should not be too difficult, ie, homesteads however defined do not come and go every day.  Aoziwe (talk) 06:32, 26 February 2019 (UTC)
 * Sorry for the belated comment, but pretty much a homestead is a home on a farm (so there are many of them out there). Historic is whatever age we want it to define it to be (but, no matter what cut-off we choose, there's going to be many of them out there). As we are discussing private properties, generally there isn't going to a way to even know about most of them, or get information on them. As cities expand, homesteads often end up in the suburbs as houses with their land subdivided into suburban lots, and often the homestead itself gets used for other purposes. Do we include these too? The inclusion criteria for this list is vague. We have no way to judge its completeness. To me, this is a bad sort of list to have. I think having an article or being on a heritage register (that would quality for notability). Both of these are easily decidable and completeness is attainable. A third criteria could be a couple of citations that would meet WP:GNG which is sort-of decidable but completeness is unknown. It's easy to start such list articles, it's hard to complete and maintain them. Maintenance is a massive issue for Wikipedia generally; we have neither the tools nor the manpower. So I tend to be a bit discouraging towards list articles that are likely to be problematic long term for these kind of reasons. Kerry (talk) 04:41, 25 April 2019 (UTC)

Wikimedia Australia Community Conference, Sydney, 15 June 2019
For more information, please see Wikimedia Australia Community Conference, Sydney, 15 June 2019 via  Kerry (talk) 10:18, 15 May 2019 (UTC)

New South Wales Heritage database
This discussion was moved from User talk:Rangasyd on 21 October 2019. Rangasyd (talk) 11:05, 21 October 2019 (UTC)
 * I can sort-of produce Wikipedia articles from these. See User:Kerry Raymond/sandbox for St Johns at Wollombi (note it looks scrappy because the NSW HD entry is scrappy). The tool I built was for the purpose of generating additional text for NSW SHR entries that were minimal but there was better content for that site in one of the other NSW HD entries. That is, the goal was to produce history, description and significance content rather than infoboxes, categories etc. Having said that, I think the main thing my tool does not do is state which heritage register it is coming from (in this case it is the Hunter Heritage Regional Environmental Plan -- whatever that is) as that wasn't needed for the NSW SHR purpose. I probably could extract the name of the heritage register from the NSW HD if needed. I assume my photo extracting tool would work too again with a little tweaking (noting I have yet to extract and upload about half the photos from the NSW SHR which is probably more of a priority for me). The main question is whether the arbitrary NSW HD entries are notable under WP:GNG. If they are, then tell me the database ID (which appears in the NSW HD URL) and I can email you the draft article. Kerry (talk) 00:56, 15 October 2019 (UTC)
 * Thanks for that. Initially I thought we'd missed one from the NSW SHR. Then I realised it was from the NSW HD. I've created the article at St John the Evangelist Church, Wollombi using what was in your sandbox as the base. Thanks so much. I'm working through Edmund Blacket's buildings. I think I will move them out of his article and into a operate list. As I was going through the articles I realised that quite a few did not have Wikipedia articles... it was only on closer inspection that I rebased that they were not in the NSHSHR. So thanks; but no need for more at present. I'll let you know if more are required. By the way, I used Infobox church instead of Infobox historic site. I can't recall if a code was ever created for NSW HD. Can you... and what is the page where you request a code be created? My memory is fading!! Rangasyd (talk) 15:49, 15 October 2019 (UTC)
 * Template talk:Designation is the place to request new codes/colour schemes set up. Kerry (talk) 20:56, 15 October 2019 (UTC)
 * Well well well, take a look at Kerry (talk) 21:55, 17 October 2019 (UTC)
 * Also I agree with you that the colours should be different to NSW SHR. Kerry (talk) 21:57, 17 October 2019 (UTC)
 * This would be great to do for the country railway stations especially, nearly all of which had NSWSHR articles which were balls and great content elsewhere in the NSWHD. The Drover&#39;s Wife (talk) 00:50, 18 October 2019 (UTC)
 * Well, if you just want content for NSW SHR articles, I can do that now. With more work on the code, I can possibly do a bit better in terms of producing stand-alone articles for things not in the NSW SHR but you feel can pass GNG. Now that the NSW HD is a designation for infobox historic site, I can produce a proper infobox for the articles, although it's a bit of a cheat as we probably should also say the precise register it is on. I am thinking along the lines of a lede that says "It was added to the Hunter Heritage Regional Environmental Plan on 18 October 2019". That would be honest about the heritage listing but not require us to have designations or individual articles about all these little heritage registers (I guess we could do redirects for them and let the New South Wales Heritage Database article explain the exciting variety of HRs it contains). Kerry (talk) 03:31, 18 October 2019 (UTC)
 * As you might recall that in my NSW SHR project spreadsheet, there is that field called "issues" that the two of you occasionally filled in noting the lack of content in a NSW SHR article. I can easily give you that list of articles as sites you might like to investigate in the NSW HD for better content in one of the other registers. If you find something useful, send me the database ID and I'll send you the generated material. Of course if you are both interested in doing it, we probably need to divide the set by LGA (as we did before) or something else so you don't waste time working on the same thing. While not every article needing additional content may have been flagged in the spreadsheet, there are 121 currently so flagged so they would be a good place to start. Kerry (talk) 03:42, 18 October 2019 (UTC)
 * I'm happy to make a start on the "issues". So, if we use St John the Evangelist Church, Wollombi as the example for applying the NSW HD template, I suggest that in the infobox we use and, hence,  . In the lede/body of the article we simply stick to "... added to the New South Wales Heritage Database on 18 October 2019". Or Kerry, do you think we should proceed as you've proposed above? Should this discussion be moved to another page? Rangasyd (talk) 10:15, 18 October 2019 (UTC)
 * You are right we should probably be talking on another page. I think your proposal is OK for me though. Kerry (talk) 02:39, 19 October 2019 (UTC)