Template talk:Infobox Australian place/Archive 8

Archive 5

Archive 6

Archive 7

Archive 8

Archive 9

Coordinates in infoboxes

per Wikipedia:Coordinates in infoboxes, I had enabled |coordinates= as an alternative to |latd=, |latm=, |lats=, |longd=, |longm=, and |longs= for specifying the coordinates for the location map. AussieLegend, reverted the edit because this enabled the location map in many articles not needed a map. how about if we add a parameter |auto_location_map=y? the template would still function the same as before, but if you (a) have |coordinates= specified and (b) |auto_location_map=y then it will pull enable the automatic location map using the value in the |coordinates= parameter? Frietjes (talk) 14:33, 22 February 2017 (UTC)

I the code that I had before in the sandbox, and see no differences in the testcases. could someone add a testcase that's breaking with the proposed new syntax? Frietjes (talk) 14:41, 22 February 2017 (UTC)

The change primarily affects LGA articles (see Category:Local government areas of Australia), although it's not limited to these. Preference has always been for these articles to use maps showing the actual LGA, and when the infobox was converted in May 2013, it was necessary to retain |coordinates= so that we could ensure that the locator map didn't display, as it was redundant to the map image and just served to bloat the infobox. I've added a couple of examples to the testcases page. --AussieLegend (✉) 16:05, 22 February 2017 (UTC)

AussieLegend, if (a) all LGAs should not have the location map, and (b) everything else should if there are coordinates, then we can just key off of the value of |type=. I added something like that to the sandbox, but with some exceptions. if that doesn't work, another idea would be that we say by default (a) LGAs don't have a location map and (b) everything else does have a location map, but then have a parameter to "force" the appearance or suppression of the location map. Frietjes (talk) 18:33, 22 February 2017 (UTC)

A third, more complicated idea, would be to scan the image parameters to see if there is a map in there. that would work, assuming all the LGA maps follow the same pattern. Frietjes (talk) 18:36, 22 February 2017 (UTC)

Preference was for maps images but not all LGAs have a map image. Some therefore use the locator map function. It would be necessary to identify those that do have an image. Offhand I don't know how many images that would be. LGAs aren't the only articles that use map images, there are several (again I don't know how many) that do. Some of those that do are cadastral units. It should also be noted that the map images usually use one of the image parameters, not |alternative_location_map= or |use_lga_map=. Unfortunately, the discussion that deprecated use of coordinate parameters never took into account the 10,000+ articles that use this infobox, and it's opened a huge can of worms. --AussieLegend (✉) 18:45, 22 February 2017 (UTC)

AussieLegend, we can always use the extremely safe option which is (1) if |latd= and |longd= are set, then we show the location map, (2) if |coordinates= is set _and_ |auto_location_map= is set, then show the location map with the coordinates extracted from the coordinates parameter. that would ensure that the old behaviour works the same, but we can transition to the preferred syntax by adding |auto_location_map=y as the transclusions are converted. Frietjes (talk) 19:31, 22 February 2017 (UTC)

to get a better idea of the number of (1) LGAs with enabled location maps, and (2) non-LGAs with suppressed location maps, I have added some temporary tracking. I am interested to know how many articles would need to be handled for the proposed default LGA = no location map / non-LGA = location map to work. Frietjes (talk) 20:21, 22 February 2017 (UTC)

Is there wider experience of infoboxes with coordinates that optionally do not display a location map? As this is a standardisation process, we should not establish a variation just because we are not aware of the standard. It looks like most of our LGAs have maps as members of the first-level subcategories of commons:Category:Maps of local government areas in Australia that are used on articles about the LGA (but each state has its own style). The coverage of location maps is good for Victoria, but patchy or non-existent elsewhere.

Somerset is a Featured Article about a British LGA. It uses a custom locator map like our LGAs, and the coordinates are not in the infobox, they are at the bottom of the page source, with |display=title so perhaps the solution we need is to move the coordinates out of the infobox if we do not want the automated map. We have no way of telling how helpful it is to our readers to find a point coordinate in the infobox for an area, when it is also displayed in the title area of the page. --Scott Davis ^Talk 23:16, 22 February 2017 (UTC)

Scott Davis, the other widely-used infobox template which automatically adds a location map is Template:Infobox UK place. for that template, you set |map_type=none to suppress the map, but otherwise, one is automatically added if coordinates are found. we could have a bot add |map_type=none to all the cases currently suppressing the map, then enable the |coordinates= to be used for the location map if |map_type= is not set to none. Frietjes (talk) 23:31, 22 February 2017 (UTC)

It looks to me like we should definitely suppress the generated locator map for LGAs and support the same parameters as {{Infobox UK place}}. I'm less-convinced about suppressing maps just because they are not currently displayed, without confirming why (e.g. uses a custom map, editor didn't know how or expected {{coord}} to be parsed soon). The longer-term solution is likely to include making a better set of smaller-area locator maps (LGAs, urban areas, densely populated regions), and possibly supporting the # notation to easily allow the reader to flick between a state map and an LGA/city map, so maybe for now the bot edit that explicitly suppresses maps is enough for regular editors to notice and remove the suppression if appropriate. --Scott Davis ^Talk 00:39, 23 February 2017 (UTC)

The longer-term solution is likely to include making a better set of smaller-area locator maps - While I'd agree, there has been an ongoing problem with getting these generated. Surprisingly, when some of the LGA mergers in NSW went ahead, new maps for the LGAs were created, but no others were. I had the same problem when rewriting {{Infobox Australian road}}. We had so few road maps that I had to include locator map functionality to identify end points of roads. However, "small" roads, like Hunter Street (Newcastle), don't really benefit. --AussieLegend (✉) 05:39, 23 February 2017 (UTC)

I made some of the new NSW LGA maps. They aren't exactly "right", as I made them by taking one of the old LGA maps and colouring in the other parties to each merger. They are correct in the location of the new LGA, but they have too many boundaries in the rest of the state as I did not attempt to remove all the other boundaries that had ceased to exist.

I have a new computer since I last tried to make maps, so I might be able to set things up and do some more if I think I have time. This conversation is at least prompting me to think about it again. I like the little insets that the Victorian ones have to help the reader to locate the detailed map. --Scott Davis ^Talk 12:20, 23 February 2017 (UTC)

Proposed solution

The tracking categories are basically filled now, and it looks like we have about 4 LGAs using a location map, and about 855 non-LGAs with coordinates and no location map. In total, there are about 1050 places with coordinates and no location map (about 10 percent of the transclusions). following the suggestion above to try to make the syntax similar to {{infobox UK place}} where it makes sense, I propose

• Add a new parameter, |map_type=, with the following properties
  • if |map_type=nomap then suppress the location map
  • if |map_type=Australia then use the national map (this would eventually replace |force_national_map=y)
  • if |map_type=lga then use the location map specified by |lga= (this would eventually replace |use_lga_map=y)
  • if |map_type=<blank> and |type=lga then suppress the location map
  • if |map_type=auto and |type=lga then show the automatically determined location map
  • if |map_type=<blank> and |type=city or town or suburb or ... then show the automatically determined location map
  • if |map_type=auto and |type=city or town or suburb or .. then show the automatically determined location map
  • if |map_type=something else then show the location map named something else (this would eventually replace |alternative_location_map=something else)

in the short term, this would mean adding |map_type=none to the approx 855 non-LGA articles in this tracking category, and adding |map_type=auto to the approx 4 articles in this tracking category. once this is done, we can safely start using |coordinates={{coord|...}} for the coordinates in the location map. or, if we want to edit another 200 articles, we could add the |map_type=nomap to the approx 200 LGA articles as well and simplify the logic a bit more. comments, suggestions? Frietjes (talk) 14:27, 23 February 2017 (UTC)

I've been away, and just realised nobody had continued this conversation. In principle, I support the proposal.

1. |map_type=<blank> and |map_type=auto are not always the same thing, and for simplicity of use, the default should always be auto. I think that might be a distinction that only hits 4 articles anyway?
2. There's a conversation further above about being able to make the map scale switchable (LGA/City/state/country) if we had all the maps available. Would there be a way of being able to do that later once the maps are made if that became supported for at least some places? Perhaps |map_type=city,lga,state - I think I just implicitly assumed two more valid values for |map_type=: city (only valid when |city= has a value) and state (the current default).

--Scott Davis ^Talk 04:40, 28 February 2017 (UTC)

Scott Davis, if we make the default |map_type=auto for all types then we would need to add |map_type=nomap to 1050 places with coordinates and no location map, instead of just the 855 non-LGAs with coordinates and no location map. this would make the template logic easier to implement at the minor expense of adding the parameter to an additional approx 200 LGA articles. The current |use_lga_map=y still uses a location map, but uses the location map for the LGA instead of the state (see, for example, Bright, Victoria). Frietjes (talk) 14:58, 28 February 2017 (UTC)

We could define "auto" to suppress the map if |type=lga, then only the exception articles need to be modified. I guess I am proposing a modification so:

• if |map_type=nomap then suppress the location map
• if |map_type=Australia then use the national map (this would eventually replace |force_national_map=y)
• if |map_type=state then use the appropriate state map (current behaviour)
• if |map_type=lga then use the location map specified by |lga= (this would eventually replace |use_lga_map=y)
• if |map_type=city use a map appropriate for the |city= parameter (new feature) will often replace custom maps
• if |map_type=auto and |type=lga then suppress the automatically determined location map
• if |map_type=auto and |type=city or town or suburb or .. then show the automatically determined location map
• |map_type=<blank> has the same meaning as |map_type=auto
• |map_type= can accept multiple values separated by commas. They will expand to using the relevant maps separated by # so that the reader can flick between them. (new feature)

How does that modified proposal sound? I recognise I am making the template coding a little more challenging and hoping to make the template easier and more functional to use. --Scott Davis ^Talk 00:24, 1 March 2017 (UTC)

Scott Davis, if we are going to allow for |map_type=something else to work, we should use '#' as the delimiter for the multiple values to make things easier to code. the only efficient way to make this work will be to code the map selection in LUA. I can work on it this weekend. without multiple maps, I have something coded in Template:Infobox Australian place/map, but that will eventually be deleted. Frietjes (talk) 23:10, 1 March 2017 (UTC)

Frietjes I'd be OK with using # to delimit. I was thinking we would use ',' to delimit types of map |map_type=state,lga,city to translate to actual map names delimited by '#' {{Location map|Australia Queensland#Australia Queensland Toowoomba Region#Australia Queensland Toowoomba}}. Note that we cannot have a strict hierarchy of scales, as city and LGA are not always in the same order, so the order must be specified by the parameters to the infobox. The city of Toowoomba is contained in Toowoomba Region, but the City of Playford is in Adelaide.

The ability to do multiple scales supported by the infobox template may be too much of a stretch to implement at the same time as the change of coordinate styles. We can have a plan for a future enhancement but just implement single scales for now if that is significantly simpler, or does most of the complex code need to be done anyway? I have hesitated to use LGA maps much as they don't provide sufficient context, but state maps are not much help for densely populated parts of a state. --Scott Davis ^Talk 23:43, 1 March 2017 (UTC)

I don't know what happened, but I had a response typed up here the other night. Oh well. The essence of what I wrote is that we need to remember that editors currently don't have to do anything other than enter raw data into the infobox. When I converted the infobox, around 2,000+ articles had to be modified to get rid of extra coordinate strings and most of the ones that I removed were wrong. Just recently there has been an increase in the number of articles in the error categories, through simple mistakes mainly. Prior to this recent "escalation", the main issue was one experienced editor who kept adding spurious pipes for a period of time. If we start changing things too much, we're likely to see an increase in the number of errors, especially if the changes are largely redundant. For example, |map_type=state is in this category. The map is currently determined by the value of |state=, modified as necessary by |force_national_map= and |use_lga_map=, so there is really no need to introduce another parameter to modify it. |force_national_map= and |use_lga_map= are both relatively low use parameters. |force_national_map= is only supposed to be used on capital cities and places that cross borders, of which there aren't many. |use_lga_map= is under-used for reasons I've explained earlier.

The ability to do multiple scales supported by the infobox template may be too much of a stretch to implement at the same time as the change of coordinate styles. We can have a plan for a future enhancement but just implement single scales for now if that is significantly simpler - Based on the sole previous discussion regarding this, I'd agree with a later implementation. In the past we had one discussion asking for a locator map (we actually always had one but it was too hard to use) and everyone has seemed happy with the current locator map so I don't see this as a high priority at this time, not that I see forcing people to manually create coordinate strings as a high priority. --AussieLegend (✉) 03:57, 2 March 2017 (UTC)

I agree with the intent of being simple for users. |map_type=<blank> should be the same as |map_type=auto which should be pretty much the current behaviour.

I aspire to making more LGA locator maps, and possibly city locator maps when I have time, inclination, data and tools. The ability to set |map_type=lga,state would make it easier to introduce them without having to make LGA locator maps with inset location maps like the current suite of Victorian LGA maps have. I have not noticed yet if the Wikipedia Maps Department has a policy or guideline about that.

I'm not sure how important it is to deliberately suppress maps from (non-LGA) articles that presently have coordinates but no map. I suspect most of those are suburbs that would be better served by a smaller scale map (LGA or city) as there is little difference in dot location on the South Australia map for Davoren Park, Hillbank, Echunga or the NSW map for Penrith, Bondi. --Scott Davis ^Talk 00:53, 3 March 2017 (UTC)

I don't think we need to use |map_type=lga,state to introduce new maps. Adding |use_city_map= should be able to achieve that, or we probably could change the existing code to select the appropriate map automatically. There aren't a lot of city maps at the moment so it's not really an issue. City maps are something that should be taken up with the Australian roads project, as they would be useful for {{Infobox Australian road}}.

I'm not sure how important it is to deliberately suppress maps from (non-LGA) articles that presently have coordinates but no map. It's been a while, but when I checked a lot of these articles I found they just hadn't been converted to use a locator map. When the locator map was first introduced long ago, it required separate parameters. That was fixed later on but, in the meantime, a lot of articles were missed. It's really only the LGA articles, and a handful of others, that should have the locator map suppressed. --AussieLegend (✉) 06:27, 3 March 2017 (UTC)

Proposed transition code

I have created Module:Australian place map, and added it to the sandbox template. there should be no differences in the testcases. this code allows the coordinates for the location map to be passed through either |coordinates= or through |latd=/|longd= or through |latitude=/|longitude=. to suppress a location map, one can use |map_type=nomap. for non-LGAs, the default is to show the map. for LGAs, the default is to not show the map. to force an LGA to show the state location map, one can use |map_type=state. to force a non-LGA to suppress a map, one can use |map_type=nomap. before this new transition code can be rolled out, we would need a bot to add |map_type=nomap to the approx. 855 non-LGAs with coordinates and no location map, and add |map_type=state to the approx 4 articles in this tracking category. we can discuss more features later, but for now, I am hoping to come up with a solution which will allow the bot to facilitate Wikipedia:Coordinates in infoboxes. comments/suggestions? Frietjes (talk) 14:37, 5 March 2017 (UTC)

The above assumes that the 855 non-LGA pages without a map made a considered decision not to have a map. Whereas I think for most of the Qld ones at least, it wasn't a considered decision but just a consequence of using the coordinates field rather than the latd/longd fields. Personally whenever I stumble over these, I've been switching them over to latd/longd so the map is generated. I suspect the cases where a custom map has been created is a tiny minority; is there a way we can detect these and only generate the nomap when the custom map is present? I think otherwise the default should be to force the map. Kerry (talk) 00:57, 6 March 2017 (UTC)

Kerry, the category has a list of all the cases. so, I would think the first thing to do would be to go through the articles in that category and check them. of course, if we start adding a |map_type=nomap, we can always have a category to keep track of all transclusions which are suppressing the map for review later. that is, unless you can think of a pattern we could use to try to make it automatic? Frietjes (talk) 14:21, 6 March 2017 (UTC)

I agree that it is probably not that big a job to manually check them, given I have just manually checked all of the Qld ones this morning, and since I am converting a lot of them as I checked them, there will be even fewer articles in the category anyway. Kerry (talk) 02:59, 8 March 2017 (UTC)

I agree with Kerry that the map would not hurt the majority of cases. Many will be suburbs that a state map won't help a lot either, and a few will have their own maps that deliberately don't have the auto map. I know I've switched a few to have the auto map, and even move coordinates from bottom of page to infobox for some in anticipation of an auto map soon. Something removes the coord if I move them up then add the old parameters as well. I doubt it's a show stopper either way, especially if they can be tracked for manual Review. --Scott Davis ^Talk 13:36, 7 March 2017 (UTC)

Well, I have manually reviewed every Queensland article in this tracking category. I found precisely one article Rolleston, Queensland among many Qld articles that had a custom map in the image field and not a generated map. The rest without exception simply used the coords without any custom map and I could see no reason why they should not have a generated map (and am in the process of converting them to do so). It is probably worth noting that the articles involved were not completely random, but were often tightly clustered (e.g. suburbs of Toowoomba, suburbs of the Sunshine Coast and national parks tended to use the coords rather than latd etc), suggesting a common authorship. So unless someone is aware of any known pockets of hostility to the use of generated maps, I would be inclined to make that the default. Even if we added a generated map to Rolleston, it would hardly do any great harm to the reader. Kerry (talk) 02:57, 8 March 2017 (UTC)

I had wondered by the earliest alphabetic ones were all generating maps, but I see now that User:Frietjes has been hard at work! Thanks for that! Kerry (talk) 04:49, 8 March 2017 (UTC)

I note that File:Rolleston location map in Queensland.PNG dates to 2005 and its use in the article pre-dates the infobox, and certainly the locator map. As I said earlier, articles not using the locator map are mostly just articles that were never converted to use the locator map. --AussieLegend (✉) 20:24, 8 March 2017 (UTC)

I have added |map_type=nomap to all the non-LGA articles with coordinates, but without location maps. those pages are now listed in Category:Pages using infobox Australian place with an explicitly-suppressed location map. all the other non-LGA articles either have no coordinates, or have location maps. if Rolleston, Queensland would be better using a location map, please feel free to change the map image to a location map. Frietjes (talk) 20:33, 8 March 2017 (UTC)

I've been bold and changed the map at Rolleston, Queensland to use the auto map, put the photo in the infobox, and noted that the 2006 population is not in the 2011 citation. --Scott Davis ^Talk 08:02, 10 March 2017 (UTC)

I have deployed the new code. there should be no visible change in the articles with (1) all the articles that had a location map before will still have a location map, and (2) all the articles that didn't have a location map before will still have no location map. with this change, switching from |latd= and |longd= to |coordinates= will not impact the appearance (or lack of appearance) of the location map. if there are any problems, please let me know so we can add those cases to the testcases and debug. thank you. Frietjes (talk) 15:32, 10 March 2017 (UTC)

2016 Census data released

Data for the 2016 census has now been released. {{Census 2016 AUS}} is now available for citing populations. Please let me know if you find any errors when using the template. --AussieLegend (✉) 07:05, 27 June 2017 (UTC)

Populations from wikidata

Hi all. A few of us been working on an effort to get Australian place articles linked to ABS id codes in Wikidata (so far for the 2011 IDs, see wikidata:Property:P4014). I also intend to link to 2016 codes as soon as possible (wikidata:Wikidata:Property_proposal/Australian_Statistical_Geography_2016_ID). One this is up and running, it will be fairly easy to import population data in bulk (with some judicious rules for places that have multiple identifiers depending on how broadly the region around them is specified). At that stage, it would make sense to draw this data directly from Wikidata to this template (see Module:Wikidata for how to do this), especially if a population figure has not been set, or is old. This could save significant time updating to the 2016 census figures, especially since 10 years down the track we still have thousands of 2006 referenced populations. Does anyone here see an issue with this, or have suggestions for the implementation? Is anyone able and willing to improve this template to make a population call when no pop= field is specified? --99of9 (talk) 07:40, 23 June 2017 (UTC)

I see one problem with this: the neverending issue of the ABS's weird and wacky census districts. A couple of editors have designed and been implementing some excellent ways of explaining in individual articles what ABS population data actually refers to in the (usual) case when it's nonsensical without context (see, for example, Antechamber Bay, South Australia). I'd be concerned that turning this into an automated field would inhibit workarounds like this - when, if anything, it's as crucial as updating the figures (a figure that claims that locality has a population of 319 is manifestly and quite significantly wrong without explanatory context, and like mistakes are everywhere in our geographical coverage). The Drover's Wife (talk) 08:41, 23 June 2017 (UTC)

Yes, this is exactly the kind of issue we need to address. Firstly, and most importantly, I think we should implement it such that if the template fields are populated, that is what is used - so we can always override any automatic figures. However, I believe we can detect the need for explanations like this. In your example, a better census link is GL_SA30 (which I've now added in wikidata:Q24191058), because the census page for the GL explains the issue itself. And any time we only have a GL ID, it is going to be inadequate to draw the population automatically. In many other situations such as wikidata:Q649969, there will be multiple associated census areas (you can see 3 populated already), and we will need some rules to determine which populations to draw/deprecate/promote (I believe the en-wiki article Ulladulla, New South Wales made a bad choice on this one). --99of9 (talk) 09:31, 23 June 2017 (UTC)

The problem we have is that the ABS's weird and wacky census districts are so far out as to be completely useless in some cases. As an example with maps, Ferodale is a large suburb near me, but the CCD for the suburb only includes a handful of people living around Grahamstown Dam. (map) Most of the people living in the suburb, which is primarily rural blocks, are included in the population for Swan Bay.[1] Also included in Swan Bay's data is Oyster Cove and part of Medowie. The ABS data is pretty much useless and very misleading in a lot of locations. Calling Wikidata for this is going to introduce a lot of invalid data. Articles like Oyster Cove, if created, would have to have |pop= empty so we'd need to have some way to not call Wikidata. The population data would need to be validated before calling Wikidata, so you may as well enter it manually. You can't just automatically call it. --AussieLegend (✉) 09:57, 23 June 2017 (UTC)

Yeah, you've both highlighted some further issues with this - in the case where there are multiple associated census areas (relating to our article as we define it), there's been many cases where we should be citing multiple figures and don't. Having thought on it a bit more, I feel like the only way automation would be useful in this way would be if there was a way to only do so where the SSC and GL boundaries are exactly aligned - but I have no idea if that's even possible or how many articles it would pick up. I wonder if a better solution might be a huge mass manual blitz with as many editros as possible trying to kill all these issues in one swoop - not my usual thing, but I'd definitely have a go. The Drover's Wife (talk) 10:15, 23 June 2017 (UTC)

If Oyster Cove is made with "pop= " but wikidata doesn't have a value on that object (because it only has a GL_NSW code, which we know do not have populations), then the autocall will just fail to come up with anything, which is fine in that case. I agree that Ferodale is bad, but the autocall would do no worse than we can currently do (i.e. displaying the population=7), as long as we leave the note on it. We may even be able to detect more cases that need notes, by some scheme such as TDW's area matching.--99of9 (talk) 12:16, 23 June 2017 (UTC)

Of course, we then get the situation with Swan Bay, where the population grossly inflated because it includes people from other suburbs. In reality, the populations of Swan Bay and Ferodale would be very similar. --AussieLegend (✉) 13:44, 23 June 2017 (UTC)

To some extent we are damned if we automate and get it wrong and damned if we don't as not many articles will get updated. If we create dedicated fields in the template called pop2016 (or whatever), we can use automation purely on those fields. That will then display whatever population is currently in the template *and* whatever the 2016 automation produces. The benefit of displaying both is that it is more likely to highlight when something is very wrong with the 2016 data if it looks really different to the 2011 (or 2006 or ...) that the article already has, hopefully prompting the person to investigate more carefully. If the 2016 data in the infobox of a specific article is very wrong because of the way these State Suburbs are constructed, the 2016 fields can just be removed from that infobox and we still have whatever population data was already there from earlier censuses. And if we find far too much of the automated 2016 population data is wrong, then it's an easy matter to suppress all the automated 2016 population data by editing the template to suppress the display of any 2016 data without harming any earlier data in the infobox, while we think if there is a way to improve the automation. I think this gives us both an individual way and a en-mass way to "back out" any problems with automated 2016 data without harming any earlier manually-entered population data in the infobox. If/when the rollout of the 2016 automated approach can be declared successful, we could then consider having the template suppress other pop fields if there is a 2016 field present, but let's leave that conversation until we are confident of victory on the 2016 automated data. I'd like to go this way as I think it meets the need to "firstly, do no harm". 03:26, 25 June 2017 (UTC)

I think that might be a better solution than just updating it regardless, but I'd really like to see us bite the bullet and try to fix this problem nationally - wildly inaccurate population figures are something we've been flagging for years, and I feel like all the information needing updating anyway is a good chance to try to all pitch in to do it. The Drover's Wife (talk) 04:36, 25 June 2017 (UTC)

If we were to automate we'd have to have a bot go through and add |pop2016= to every article, but we'd also have to include a disclaimer along the lines of "this population figure may be absolute crap for this location. Don't trust it until it is verified." That means somebody manually going through every article. I have asked several times for people to go through articles and update population figures but nobody seems interested. --AussieLegend (✉) 04:51, 25 June 2017 (UTC)

According to this ABS publication (see "How well do SSCs represent the Gazetted Locality of the same name?") there are confidence levels. Assuming we can get our hands on these confidence levels and into Wikidata along with everything else, we can presumably set our threshold at which we will not use SSC population data. Perhaps we start by doing the initial rollout only on very high confidence levels and see how that goes. Then we inspect manually some random samples at the next level of confidence and, if we are comfortable, repeat the process at the next level of confidence, and so on until we feel the confidence is too low. The thing to bear in mind here is that we are faced with a choice of using ABS census data or not reporting population data. It's not as if we have any real alternative sources of population data. Think about it too from the reader perspective. Does it matter to the reader if there are 1000 or 1001 or 1100 or 2000 people living in a suburb? No, because any data we provide is somewhere between months and years out of date anyway (babies are born, people move, people die). What does matter is if there are 10, 100, 1000, 10K, 100K or 1M people because it give the reader a sense of whether we discussing a cluster of a few houses where everyone knows your secrets through to a major metropolis with unaffordable housing and traffic jams. If ABS regard confidence of 75% to less than 88% as acceptable, I think we should too. They are supposed to be the experts. Kerry (talk) 05:34, 25 June 2017 (UTC)

Very interesting. I'd like to see this trialled (for example for a specific LGA) so we could get an idea of how useful those confidence levels are going to be. The Drover's Wife (talk) 09:01, 25 June 2017 (UTC)

Does it matter to the reader if there are 1000 or 1001 or 1100 or 2000 people living in a suburb? - The problem is, some of the population figures are blatantly wrong. Ferodale is an example. It's not a case of 1 or 2 people different, in the case of Ferodale it's several hundred out of a population of several hundred. Going to extremes, the ABS definition of Sydney includes the little hamlet of Glen Alice, population significantly less than 288, listed as "Glen Davis" by the ABS, 128km as the crow flies and 223km by road from the Sydney CBD.

babies are born, people move, people die - Surprisingly though, this generally doesn't make a great deal of difference to the population figures.

These are just things of which we have to be aware. --AussieLegend (✉) 11:13, 25 June 2017 (UTC)

I'm not sure what you're saying about Glen Davis being "included" in Sydney... the current Sydney wikidata item includes associations with three different levels of definition of "Sydney", of which only "greater Sydney" includes that town. [2][3][4]. Again, whenever there are multiple associations, we can make rules about which we grab populations from when importing to wikidata. In Perth they decided to make a whole separate page about Greater Perth.--99of9 (talk) 01:00, 26 June 2017 (UTC)

This is the problem that I am *hoping* that the confidence levels will solve and we won't add something that is wildly wrong. It will be interesting to check the data and confidence levels for places like Ferodale where we know there have been problems in the past. I guess we start with checking the confidence levels to see how much confidence we have in the confidence levels. Kerry (talk) 23:01, 25 June 2017 (UTC)

How would we go about doing this? I am very bad at such things. The Drover's Wife (talk) 00:16, 26 June 2017 (UTC)

Yes Kerry, this is a good idea. I will try to bring the confidence level into wikidata at the same time as the population. Or if that is not possible, only bring in the populations with a good confidence level. I'm not a fan of the pop2016 idea, because that adds a whole lot of unnecessary bot-work when even if the idea of displaying them side by side is preferred, the template can call all the extra info from wikidata directly. --99of9 (talk) 00:50, 26 June 2017 (UTC)

We do it with our beady little eyes, comparing the maps displayed in ABS Quick Stats that we currently use to (or whatever replaces it) for a random selection of places. If the relationship of the maps is very close (as I would expect for developed suburbs etc), then the confidence should be high. So for example, here is the map displayed for Chapel Hill, Qld as a GL, which gets mapped onto Chapel Hill Qld SSC and we see that the boundaries (red and blue) are "pretty close" so I would expect that the ABS would consider this a high confidence situation where the SSC population is probably going to be pretty close to the GL. Now Chapel Hill is a well-developed "middle" suburb. But if I look at something more rural, e.g.England Creek as a GL which maps onto Wivenhoe Pocket SSC, you see maps that looks much more worrying, not at all close. I would imagine that the ABS onfidence level should be much lower as pretty clearly Wivenoe Pocket SSC isn't a useful approximation for the England Creek GL. By doing this with random places, we see if our confidence is matched by the ABS confidence or not. Kerry (talk) 01:33, 26 June 2017 (UTC)

99of9, yes, I was assuming using some kind of bot to add "| pop2016 = Absolute Value" or "| pop2016 = CallToWikiData" to the infobox instances in selected articles (where we have higher confidence). But you are right, if we can do the confidence level checking and wikidata call within the template definition (my knowledge of the expressive capability of the Template language is limited). But what do we do with a specific instance where we think the automation is getting it wrong (but is getting it right in general)? How we do "turn off" the population display in that instance but allow the others to happen? Introduce a property "dontdisplay2016pop"? Also any automation faces the challenge of having to match or modify a Wikipedia article title (which may have disambiguation involved) into the name the ABS uses. That's a heck of a lot of "smarts" to embed into the template definition rather than generate a exception file out of a bot which we scan manually as fix as needed. Just a concrete example, we have West End, Queensland and West End, Queensland (Townsville) which the ABS calls "West End (Brisbane - Qld)" and "West End (Townsville - Qld)". Our titles are not exactly in 1-to-1 correspondence. Using the bot assumption, I was assuming we'd generate as many as we could and report the exceptions. Any exception with a high population is almost certainly going to have a Wikipedia article so it's a matter of a person finding it and updating it manually. Exceptions with low populations are far less likely to have Wikipedia articles so we might not put the manual energy into them. The more I think about this, I am thinking maybe a bot working straight off the ABS data (not via Wikidata) may be easier. Use automation to add the higher confidence articles with easily matched names in the ABS with explict pop2016 fields with explicit values, give the name mismatch exceptions to humans to resolve manually. Do nothing with the low confidence data because our experience shows us it's next to useless. The devil is always in the detail. I note all of these conversations are about the infobox; we have never discussed the population data that often appears in the lede para or elsewhere in the article. I don't see any smart solution to that at the moment. Kerry (talk) 01:33, 26 June 2017 (UTC)

Kerry, have you seen Wikipedia:Australian_Wikipedians'_notice_board#2016_Census_data? We are doing the mapping between wiki articles and ABS IDs right now. We've done around 7900 already, but I will now go ahead and add the West Ends you just mentioned. --99of9 (talk) 02:16, 26 June 2017 (UTC)

I hadn't realised the conversation was split in two places. Kerry (talk) 02:45, 26 June 2017 (UTC)

Sorry about that. This conversation is an offshoot. That effort is going ahead regardless, and over here on the template we may be in a position to make use of it. @Kerry Raymond: if you or anyone else wants to help with some mind-numbing disambiguation, we'd love to have you join in at MixNMatch --99of9 (talk) 02:57, 26 June 2017 (UTC)

@99of9: Gosh, you know how to show a girl a good time! :-) I'm happy to help, but can you point me at the instructions of what I am supposed to do. I took a look at the site but I didn't quite figure out what I was supposed to be doing, presumably something with the unmatched. And, is there a way I can limit myself to Qld, as my Qld geography and knowledge of Qld Wikipedia is quite good, not so for other states, so I would expect to be more productive with Qld data, but hey I will do my best if it's not possible. Kerry (talk) 04:49, 26 June 2017 (UTC)

It's OK. I think I figured out the barrier was OAuth so I signed my life away. Kerry (talk) 05:01, 26 June 2017 (UTC)

While the discussion around potential mismatches between SSCs and localities is worthwhile, in practice the census data will be used by editors regardless of concerns raised here. It is already happening. As far as this template is concerned, the discussions should be about whether this process is better automated or left to random editors. Automation through this template and Wikidata will see a more consistent and systematic approach to population. Unless someone has a more accurate, reliable and authoritative source for populations of localities - that doesn't involve OR - the census data is what we should be using. -- Mattinbgn (talk) 04:26, 27 June 2017 (UTC)

It's every bit the question because having it uniformly wrong is not, in fact, definitively better than doing it piecemeal. And it's not a matter of whether the census data is "what we should be using" - rather, that we use it in a way that explains what figures are actually being referred to, rather than posting population claims that are literal fiction. A figure with a note that explains that the figure used is really the population for ten towns is accurate; posting it in as if the population of that town is ten or twenty times what it is because you can't be bothered explaining otherwise is just wilfully inserting fiction into the project. The Drover's Wife (talk) 04:49, 27 June 2017 (UTC)

A quick look at the released data shows some improvement over past censuses. For example, the data for Raymond Terrace now covers almost all of the suburb. It includes some vacant, never to be built on land at the south of the suburb but misses two properties at the northern part of the suburb. Ferodale is now shown correctly with the population up from 7 to 83 (I thought it would be higher). Swan Bay is also shown accurately, with the population down from 437 in 2011 to 316 in 2016. However, there is still a need for editors to verify the data and provide notes in articles where necessary. In some cases it may be necessary to remove notes. If we just blindly pull data from Wikidata without manual verification, there are going to be a lot of articles with unexplained errors and anomalies. It would be nice to automatically populate articles with population data, but we would need to clearly identify articles that have not been manually verified, both in the articles (which might fall afoul of WP:NODISCLAIMERS) and by inclusion in a category that would identify articles needing verification. Category:Australian place articles requiring manual verification of population data perhaps? --AussieLegend (✉) 08:17, 27 June 2017 (UTC)

The Wikidata id-matching already has one significant tangible benefit. This category: Category:Australian_Statistical_Geography_Standard_2011_ID_different_from_Wikidata is down from about 130 items, but is still pretty full. In many cases the current code on Wikipedia is wrong, sometimes left over from 2006 when the template was converted to 2011 example. If anyone wants to help work through them, it would be appreciated. I've started at the bottom of the list, and am working my way up. --99of9 (talk) 12:12, 5 July 2017 (UTC)

I'm not sure I understand. I picked one that was known to me, Hinton, New South Wales and there is nothing wrong. Am I to assume it is because the article uses GL_NSW1934 instead of SSC11117? If so, there is no error. The Gazetted Locality link was deliberately chosen to identify the correct map for the suburb, as Quickstats doesn't cover the actual suburb, just a part of it. The same applies for Kooragang. For other places, the UCL or SA2 area more closely matches the gazetted area than the SSC. This was a big problem in 2011. I can't see any error with Glen Oak, New South Wales. --AussieLegend (✉) 14:29, 5 July 2017 (UTC)

Pinging @Kerry Raymond:, who has been "fixing" the citations, since he's obviously not seen this. --AussieLegend (✉) 17:39, 5 July 2017 (UTC)

Yes, in that kind of case it's better to add the GL as an option to Wikidata. Then the category goes away next time the en-wiki page is edited. I've now done this for Hinton. --99of9 (talk) 22:17, 5 July 2017 (UTC)

@Kerry Raymond: did ask me on my talk page how to do this, but I had just gone to bed, sorry. --99of9 (talk) 22:20, 5 July 2017 (UTC)

Country

I just noticed that the infobox doesn't show "Country: Australia" anywhere. This had been discussed before in December 2011, and User:AussieLegend added it in July 2013, but removed it again in November that year, citing "temporarily suppressing 'Australia' in the heading, as there seems to be some opposition to it." of which I can't find any discussion. Other localised place infoboxes emit a country, as does of course the generic {{Infobox settlement}}. I suggest this one should, too. -- Michael Bednarek (talk) 11:15, 18 November 2017 (UTC)

Sounds reasonable to me. Frietjes (talk) 13:19, 18 November 2017 (UTC)

Four years ago is a long time so I don't remember the specifics but I do remember comments from several editors about removing it, which is why it was removed. I haven't checked the WP:AWNB archives but it's likely the comments were there if not here. --AussieLegend (✉) 14:20, 18 November 2017 (UTC)

AussieLegend, the thread was on your talk page. related is this thread. Frietjes (talk) 14:39, 18 November 2017 (UTC)

FWIW, it does not appear to be the normal practice to display the country name in Template:Infobox UK place nor USA articles which appear to use Template:Infobox settlement rather than anything more specific. For myself, I think it is important to mention Australia in the first sentence but I am less concerned about the infobox. As a practical consideration, our infobox is fairly bulky already. Kerry (talk) 23:01, 18 November 2017 (UTC)

Kerry Raymond, here is a UK place: Abercarn, you will see United Kingdom in the infobox. please provide an example using Template:Infobox UK place without listing United Kingdom in the infobox? as for USA articles using Template:Infobox settlement, please provide an example where "United States" or "USA" is not listed in the infobox, the only ones that I have found (e.g., San Diego) list the country in the infobox. Frietjes (talk) 13:52, 19 November 2017 (UTC)

I was specifically talking about the heading in the infobox. And random places I checked were Preston, Lancashire, Ramsgate, Wells, Somerset for the UK. You are right that it does appear lower down in UK infobox but for me that's on the 2nd or in some cases 3rd screen depending on the length of the infobox (and I have no idea where it appears in some of the mobile renderings). But if it's not visible on the first screen (ideally in the heading of the infobox), it's pretty near useless for practical purposes. When a person lands on a page, they need to know it's the correct page, which is why we have the hatnotes for disambiguation at the top and why we try to ensure that the first sentence clearly identifies the topic. So when I load Abercarn (a place I do not know), all I see of the infobox is down to the map caption "Abercarn shown within Caerphilly" (and I've never heard of Caerphilly and the map displayed doesn't help me either, although I do recognise the inset map of Wales as being Wales) and all I see of the text is down to the heading "Local government". The lede para tells me it is in Caephilly (a name I don't know), then Wales (which I do know) and then mentions Monmountshire (which I do know). So I am now "oriented", I know where Abercarn is more-or-less. Having mentioned UK in the first sentence would have been better in my opinion, as Wales may not be well known world-wide (population 3 million, so less than most Australian states). Now I do know where San Diego is (so it's not a real test for me) but if I didn't, the information it is in the USA is in the lede sentence (but not visible in the infobox to 2nd screen). I think if you include the country in the infobox, it needs to be in the heading of the infobox. If you have to get to the 2nd/3rd screen to figure out which country you are in, you must have already worked out the country from the text or you have a lot more time on your hands than I do. While I don't know of studies relating to user behaviour on Wikipedia pages, search engines are well-studied and we know the vast majority of people do not look at the 2nd page (which is why we have an industry of SEOs - search engine optimisers - who you pay to get your website onto that first screen of search results). By all means include the country but put it in the header where it might do some good. Kerry (talk) 22:04, 19 November 2017 (UTC)

List item tags

This template contains unmatched <li> tags. They currently seems harmless but may become problematic when Tidy is removed: see WP:VPT#Tech_News: 2018-11. As many of the tags are in conditional code, would an expert like to consider adding matching </li> tags? Thanks, Certes (talk) 13:45, 18 March 2018 (UTC)

Interest in using Infobox mapframe?

Good morning. Is there interest in making mapframe maps an option for this infobox?

{{Infobox Australian road}} has been able to draw maps using {{infobox mapframe}} since May. It gets coordinates from wikidata and the map and feature from OpenStreetMap. Some of these look better than the locator map, although there are still some issues too (eg scale is defaulted or drawn from Wikipedia length parameter, centre is sometimes one end of the feature with the other end off-map). Ideally it should be possible to turn on/off each kind of map so that infoboxes on short articles don't end up with two maps. It is possible to use it now in an ad hoc way, my examples are Smithfield Plains, Waterloo Corner, Parafield, Maude. Scale could be set from the area of the feature, but we have not typically included area in place infoboxes. The maps would benefit greatly by a mass addition of wikidata IDs to OSM (or possibly OSM relations to wikidata - I haven't checked if that works). Pinging a couple of regular contributors to both infoboxes, apologies if I missed someone who should be included. @Frietjes, AussieLegend, and Evad37: --Scott Davis ^Talk 02:33, 24 September 2018 (UTC)

I'm all for it - I agree they're often more informative and it would be preferable to integrated it into the infobox rather than rely on ad-hoc solutions. The Drover's Wife (talk) 02:44, 24 September 2018 (UTC)

I'm sure we've discussed something similar to this previously, and I support the idea in principle. However, it's necessary to note that the primary focus for all encyclopaedic articles are readers who are not necessarily familiar with the subject and we therefore need to provide some "locational context" so readers know where these places are in relation to larger areas. That's why the convention has always been maps of Australia for the capital cities and state maps for other cities and towns. Suburbs have always been an issue as they really require more detail which this proposal provides. However, I don't see that maps like the one in Maude, South Australia (see above) are helpful at all. We need to ensure that we can provide a default map that provides a clear indication of where a location actually is. I'm not sure what is meant by we have not typically included area in place infoboxes. |area= does exist and is used quite a lot. Census Quickstats data includes area in downloadable spreadsheets, although this is not necessarily the same as the actual area. --AussieLegend (✉) 11:29, 24 September 2018 (UTC)

I agree the map showing a point on a bend in a road (somewhere) is not helpful. I deliberately picked somewhere less obvious as one of my examples, and it's possible that the state locator map will remain appropriate for that place, although learning more about controlling scale and centre, it might be most helpful to have the map scaled to show that Maude is along the road from Burra on one side to Morgan on the other. I have added the Wikidata ID to the boundary relation of Maude in OpenStreetMap, so it is possible the map will become more helpful when it shows a shape rather than just a point. I didn't know that the area of suburbs and localities was readily sourced through the census data and I don't think it's very common on SA place infoboxes. It might raise another problem to use that often for whether our articles/infoboxes should document specifically a built-up town or the much larger locality for which that word is the address. My other mapframe examples at the moment were outer suburbs of Adelaide with the existing maps being a local government area, Adelaide built up area (which is very tall and narrow, as that is the shape of Adelaide between the hills and the sea), and state. I do not propose that the locator map should be universally replaced or supplemented by a mapframe map, I would like to have a simple and consistent way of using it if required, switches to choose which one to use, and some style guidelines on how to choose which one to use when (which might be clear enough to implement as default choices). --Scott Davis ^Talk 14:01, 24 September 2018 (UTC)

I like having both the zoomed in mapframe and the larger context from the location map, but having both showing at the same time may be a bit much for some articles (some people like lots of maps, others don't). we had some discussion at Module talk:Location map about allowing a static image (or possibly a mapframe) integrated into the multimap selector, but we never executed the idea in code. if there is interest in having a mapframe be one of the switchable options (possibly the default), we can discuss how to make it happen at Module talk:Location map. Frietjes (talk) 14:08, 24 September 2018 (UTC)

It'd be useful to have some sort of consensus about when these should be used, too - now that it's started to take off I've already seen the issue AussieLegend raised with Maude popping up in article space on several articles, which is probably unhelpful. The Drover's Wife (talk) 04:43, 27 September 2018 (UTC)

I support this where it appears useful. By the way, the wikidata property to add an OSM relation is wikidata:Property:P402, which I've just added for Smithfield Plains. --99of9 (talk) 06:54, 27 September 2018 (UTC)

I just tried rescaling the Maude example. Does that look better to others? --99of9 (talk) 07:03, 27 September 2018 (UTC)

That's better. I'm not a fan of the box that has been drawn around it, although I know that's not you. Any outline should really be the border of the town or suburb, not an arbitrary box. I tried using |zoom= but that doesn't really help as Maude is still just a place with very little around it. By contrast, I added a mapframe to Raymond Terrace with |zoom=9 and I think it works well. The normal map shows the location within NSW and the mapframe supplies sufficient information about surrounding areas without overwhelming the prose. Zoom=7 covers too much area, zoom=8 doesn't supply enough local context and zoom=10 loses Newcastle (a major city) so the reader doesn't get a good idea of the location. Maitland and Hexham aren't identifiable enough. We do have a number of articles with two images so, if we were to use the mapframe it would be a good idea to provide extra fields in the infobox to enable the mapframe. I really see this as an option, rather than something that would replace the existing map. --AussieLegend (✉) 08:31, 27 September 2018 (UTC)

The "box drawn around Maude" is not an arbitrary box. it is the boundaries of the locality drawn from OSM. It appears that there are two ways of linking OSM and Wikidata. I think the mapframe can use either. The one I have used (because it was what I found first) is to add the Wikidata Q-number as the wikidata attribute to the relation in OSM that represents the boundary (or route for Infobox Australian road). The other way is to put P402 on the Wikidata item as 99of9 has done. Smithfield Plains already had the wikidata ID on the relation in OSM. I have just added the OSM relation ID to Waterloo Corner, and don't immediately see any difference. I know it takes a day or two for updates to OSM to flow through, so it is possible that it takes time to flow through from Wikidata as well. --Scott Davis ^Talk 09:50, 27 September 2018 (UTC)

PS If anyone is interested in helping add OSM relation IDs to wikidata items, here's your chance for NSW localities MixNMatch catalog 1228. --99of9 (talk) 02:33, 28 September 2018 (UTC)

Is the geometry from OSM supposed to be picked up by the mapframe regardless of whether the link is placed on Wikidata or OSM? It seems to take a day or two after a wikidata ID is added to an OSM relation before it appears in mapframe, but even longer if the Relation ID is put into wikidata only. Is there a bot that finds links that are only in one direction, and creates the inverse to match? --Scott Davis ^Talk 00:55, 29 September 2018 (UTC)

Map

Most of articles about cities has a topographic map of country/region in infobox, for example: from Los Angeles in America, Warsaw in Europe to Auckland in Oceania. The topographic (relief) map has appropriate colors showing the terrain and imposed boundaries of regions - looks better and has more information than yellow map. I suggest - in Infobox Australian place - replacing the map to relief map. Map has the same parameters (co-ordinates) and operating in many other projects, e.g. pl:Sydney. Subtropical-man (talk / en-2) 19:11, 5 October 2018 (UTC)

Sydney

Sydney

Sydney New South Wales
Sydney
Coordinates	33°51′54″S 151°12′34″E

This is already a documented feature of the infobox and is enabled by setting |relief=1. It isn't suitable for all articles, which is why it is an option. --AussieLegend (✉) 19:22, 5 October 2018 (UTC)

Here is the problem - it is an option. It isn't suitable for all articles? - please example(s). By the way, even if very small part of articles about Australian places has some problems with relief map (I guess it's about aesthetic issues), in this articles we can use the yellow map. I suggest - change to the relief map as displayed automatically, if any article will have a problem, can use |relief=0. Subtropical-man (talk / en-2) 19:49, 5 October 2018 (UTC)

I couldn't give you examples off the top of my head as the issue was identified over 5 years ago. The map caption is almost impossible to read in some articles, which is why it was left as an option. It's best to have the infobox working properly in all articles rather than not working in some. The point of the map is to display the location, which it does with the standard map. Displaying topography isn't an essential function and it really doesn't tell you anything. It just makes it look more pretty in some cases. If you look at the examples on this page, Sydney is much easier to see on the standard map. --AussieLegend (✉) 20:34, 5 October 2018 (UTC)

Given this template is used on thousands of Australian articles, I think any discussion about changing it should be drawn to the attention of folks at Wikipedia:WikiProject Australia and the various sub-projects as that may better tease out the pros and cons of the alternatives. FWIW, being older, I tend to prefer a less busy background for maps for ease of reading the labels, but I agree that the relief map looks prettier; however, I generally favour substance over style.Kerry (talk) 05:01, 6 October 2018 (UTC)

Indigenous names

What's the preferred way to handle this? Quite a number of places are gazetted in Queensland with Indigeneous names listed as alternative names. Currently we don't have a field in this template for either alternative names in general or Indigenous names in particular. For example Great Keppel Island is gazetted with 3 alternative/Indigenous names. Is the way I've dealt with this in the article (both the lede & the infobox) the preferred method or is there a better way? I think where we know an Indigenous name for a place, we should include it. Kerry (talk) 23:30, 23 November 2018 (UTC)

It would be good to have a field for this somewhere - it's definitely something that should be in there. The Drover's Wife (talk) 00:29, 24 November 2018 (UTC)

I looked at this a few years ago and there were actually very few places, compared to the number of uses of the infobox (now 13,082), that had gazetted indigenous names. Of those that do have a gazetted name, most are pretty much unknown (How often have you actually heard someone refer to "Wapparaburra"?). It's probably best to handle this in the lede rather than introducing additional, very rarely used fields in the infobox. If you do want to use a native name then {{native name}} should first be modified to support indigenous names. --AussieLegend (✉) 04:46, 24 November 2018 (UTC)

What is the issue {{native name}}? Is this not the result that is needed:

• test (Wiradjuri)
• test (Gamilaraay)
• test (Ngarrindjeri)
• test (Adnyamathanha)
• test (Guyani)
• test (Yankunytjatjara)
• test (Pitjantjatjara)?

Compare with Deutschland (German). — Preceding unsigned comment added by Gonnym (talk • contribs) 12:55, 26 November 2018 (UTC)

iucn_ref

Could you add the parameter "iucn_ref" to this template, so that it is like {{Infobox protected area}}? A number of the protected areas in Australia could use them:

• Boorganna Nature Reserve
• Bretti Nature Reserve
• Barrengarry Nature Reserve

--David Tornheim (talk) 01:11, 30 December 2018 (UTC)

Done I've checked all the artcles that you added references to and they seem to be working fine. Thanks for identifying this problem. --AussieLegend (✉) 09:40, 30 December 2018 (UTC)

Some protected areas have multiple IUCNs (e.g. Western Kangaroo Island Marine Park (state waters); is there anyway to include multiple designations in an infobox? Regards Cowdy001 (talk) 01:11, 9 January 2019 (UTC)

Thanks for taking care of this. Sorry for the late reply. --David Tornheim (talk) 23:24, 14 January 2019 (UTC)

generates a non-standard infobox for cities

Hi, I just noted, that this infobox generates a different html table class than the infoboxes used for most other countries' cities. Should we fix it?

<table class="infobox vcard" ...

instead of

<table class="infobox geography vcard" ...

46.235.154.83 (talk) 17:17, 4 February 2019 (UTC)

Nothing to fix. This is entirely appropriate. Nothing is mandated. --AussieLegend (✉) 17:19, 4 February 2019 (UTC)

Could you please describe why it is appropriate to use a different class for the html tables of Australien cities' infoboxes compared to almost all other cities in the world? --168.149.159.22 (talk) 08:07, 5 February 2019 (UTC)

Rainfall

The bottom of the Top End infobox says:

Annual
rainfall
1,000-1,500 mm
940.9 in

1000 mm is 39.4 inches, and 1500 mm is 59.1 inches, but 940.9 inches would make that grassland the wettest rainforest on earth. The template presumably wasn't intended for a nonstandard input like 1000-1500, but you might want it to fail more gracefully. An error message maybe, or omit the conversion, rather than such a completely wrong answer. Art LaPella (talk) 03:35, 9 February 2019 (UTC)

Direction of other regions

When this template is used for regions the direction table shows “Localities around xx (Region name)” instead of “Regions around xx (Region name)” Downsize43 (talk) 01:44, 23 June 2019 (UTC)

Mass changes to NSW articles use of this infobox

Please join the discussion at Wikipedia:Australian Wikipedians' notice board. Kerry (talk) 23:08, 29 July 2019 (UTC)

Provision for multiple states

Currently this infobox can cope with places being in multiple local government area, multiple state/federal electorates but not multiple states. We have a problem in at least one existing article Mungindi which crosses the QLD/NSW border. And there is a current proposal to merge Mingoola, Queensland and Mingoola, New South Wales (another cross-border place). Can this infobox please be expanded to have state2 in a similar way to lga2, stategov2, etc. I don't know if there are any place that would require state3 (the state "corners" come to mind and I note none of the infoboxes for those 4 articles currently use this infobox (some use a Queensland Heritage Register infoboxes as the survey marker is heritage listed) and the other has no infobox). I think state4 is geographically impossible though. Kerry (talk) 01:03, 10 November 2019 (UTC)

I actually replied at the Mingoola article while you were posting here. Yes, this is an issue but it's something that affects only a few places. While it's a bit of a hack, using the infobox twice should be okay for these limited instances. Fully populate the first with everything for the first state and only add what is necessary in the second. Modifying the infobox to accommodate two states for only a very few places is unnecessary complexity in my opinion. I did it in {{Infobox Australian road}} so it can be done, but it was absolutely necessary there. --AussieLegend (✉) 01:38, 10 November 2019 (UTC)

Traditional owners/custodians

Whenever I look up an Australian place I am interested in who the traditional owners/custodians are. Some pages include it but it's not standard in the infobox and perhaps it should be, since (unlike e.g. the multiple states question) it's relevant on every part of the continent, even though in some cases the ownership is contested or unclear. I've looked in the archive for discussion of this but couldn't see it - apologies if I missed it. (There was discussion of Indigenous names but that's a separate issue.) Spoonriver (talk) 03:26, 26 November 2019 (UTC)

If this is to be included, it should require a reference. The guidance should also indicate that the infobox field should only be populated if the prose describes the situation. As you said, there are many places that are contested or unclear, or the "boundaries" moved over time, before or after white colonisation/settlement. I'm not sure what the infobox label should be, either. "Traditional owner:" is wrong for (almost all) towns, as the town wasn't owned by/in the custody of the Aboriginal or Torres Strait Islander group, only the land that is under and around it. It might be true to say "the Kaurna were the traditional custodians of the Adelaide Plains", but it is not true to say that about every town and suburb on them. Also, the modern boundary between the Adelaide Plains and the Adelaide Hills for example may not be in the same place as the boundaries between the lands of the Kaurna, Peramangk, Ngarrindjeri, Ramindjeri, Narungga peoples, and they appear to have moved as some tribes flourished and others were affected by disease or famine.[5] That's my area, but I think I've heard similar issues for the inner southern part of Melbourne for example. --Scott Davis ^Talk 06:44, 26 November 2019 (UTC)

Agree that the field should only be populated where the info is known: that seems intuitive to me. In many places the reference would be the Registered Aboriginal Party. re the terminology, fair point, but I think 'traditional owners:' is a good starting point. Perhaps a useful analogy is a local government area or federal electorate: it's a larger area that contains whatever's within it, whether or not the name of any given part changes. re contestation, as far as I can tell a minority of places are contested, and in the majority of places traditional ownership is clear. Where it's not clear it could be left blank, or could be described as contested. My bias would be to stick to contemporary territories (e.g. RAP areas) for an infobox, just as we do for things like federal electorates even though those boundaries change over time too; I think description about changing territories over time is better suited to article copy where it's known, because it takes space to describe properly. Spoonriver (talk) 09:25, 27 November 2019 (UTC)

State electorate(s); what about upper house electorates?

Presently, the state electorate(s) field only lists the lower house electorate(s) that a place is in. Despite this, in some states the upper house has more than one electorate; in the case of one state (Tasmania) the upper house has even more than the lower house does. Shouldn't the upper house electorate(s) a place is in also be listed in the case of states where the upper house exists and isn't elected at large? Fuse809 (contribs · email · talk · uploads) 23:15, 20 December 2019 (UTC)

That's a good point. We probably should have such a field. 12:43, 21 December 2019 (UTC)

What states other than Tasmania have upper house electorates? Generally, lower house electorates are far more relevant to places on a day to day basis than the upper house so there is good reasoning behind including the lower house but I don't yet see why we need to include upper house electorates just because they exist. --AussieLegend (✉) 15:13, 21 December 2019 (UTC)

The Victorian Legislative Council and Western Australian Legislative Council both have multiple electorates. In Australia, upper houses are almost as powerful as lower houses, so I fail to see how they're less relevant. Fuse809 (contribs · email · talk · uploads) 20:25, 21 December 2019 (UTC)

I would support naming them where they exist, assuming they are easy to source accurately (I expect they are). I had not noticed their absence before as I live (and vote) in SA, which abolished multiple upper house electorates in 1975. --Scott Davis ^Talk 10:50, 22 December 2019 (UTC)

conversion to Template:Infobox settlement

Hey everyone,

Can someone please please convert this template to Template:Infobox settlement just like they have with Template:U.S. state? -- PK2 (talk) 23:27, 24 January 2020 (UTC)

This template exists precisely because it is more specific to Australian needs, not through any lack of awareness of that other template. This is not to say it cannot be discussed, but maybe you might like to articulate the benefits of making the change. Kerry (talk) 04:59, 25 January 2020 (UTC)

This actually was discussed in the past and nothing has changed since then, except that the scope of this template has expanded significantly. --AussieLegend (✉) 06:28, 25 January 2020 (UTC)

The deletion of Category:IUCN Category Ia and the effect on protected area articles using the Infobox Australian place

Hi everyone,

The following is for information purposes only.

Yesterday, I noticed that Category:IUCN Category Ia has been deleted - please refer the following for the notice of deletion Category:IUCN_Category_Ia. All of the articles that I have written about Australian protected areas with an IUCN Category Ia have been affected. For example, 'Category:IUCN Category Ia' no longer appears in the category list for Althorpe Islands Conservation Park.

I posted a message on the talk page of WikiProjects Protected areas which is the relevant project - please refer Wikipedia talk:WikiProject Protected areas#The deletion of Category:IUCN Category Ia.

Regards Cowdy001 (talk) 19:27, 16 March 2020 (UTC)

It seems rather strange that the category has been deleted when IUCN Category 1a still exists. I don't understand the deletion rationale that it is non-defining, especially when it's the highest level. --AussieLegend (✉) 20:20, 16 March 2020 (UTC)

Template-protected edit request on 31 March 2020

Hey! I think that the bullet point before the Density parameter might be a mistake. ItsPugle (talk) 03:00, 31 March 2020 (UTC)

It is a sub-entry under "Population". What do you propose instead? – Jonesey95 (talk) 05:03, 31 March 2020 (UTC)

Oh, I don't really think it's a sub-entry? I mean, like I can understand why it might be, but there's no real point as far as I can see in making it a sub point - it's the only one? Maybe just removing the dot point might be the best thing to do, otherwise it looks really... funky. ItsPugle (talk) 05:15, 31 March 2020 (UTC)

It's definitely not a mistake. It's a sub-entry of "Population" as Jonesey95 indicated. Otherwise we'd have to have "Poulation density" as a label. --AussieLegend (✉) 06:04, 31 March 2020 (UTC)

It's a subentry of "Population" and "Area", and so I agree with ItsPugle that the bullet point looks a bit out of place. It would be better to move the line "Established" above "Population", then "Area", then bulleted "Density":

Coordinates
Established
Population
Area
 • Density

This way, it looks more logical. -- Michael Bednarek (talk) 06:23, 31 March 2020 (UTC)

Maybe it's worth just calling it population density? If that's too long, just put in a few spaces as an indent, otherwise it looks like a mistake to the general public. ItsPugle (talk) 06:30, 31 March 2020 (UTC)

While it may be related to both population and area, its primary relevance is to the population figure. Putting it after area makes it look relevant to area, but not population.

otherwise it looks like a mistake to the general public. - In the almost 13.5 years that this template has existed, you're the only person who has ever complained. --AussieLegend (✉) 09:28, 31 March 2020 (UTC)

I mean, okay, so I think it's just the bullet that's making it look funky. What about my idea of an indent? ItsPugle (talk) 10:57, 31 March 2020 (UTC)

Can you please be more careful when responding? One of your responses deleted Michael Bednarek's post,[6] and you just deleted the middle of mine.[7] As for the bullet, it has been in the template for 13.5 years with no complaints so you need a much wider consensus to get it removed. --AussieLegend (✉) 11:47, 31 March 2020 (UTC)

Apologies for the issues with replying; my ISP is somewhat struggling to keep my internet on, so I'm sorry if my messages are overwriting yours - to me it honestly is that your messages aren't showing. Nonetheless, just because the template's existed for 13 and a half years doesn't mean it's not imperfect nor that there can be improvements. Not to mention that this whole conversation is about a bullet point, not the principle behind the template or the template itself, so this polarity of a community discussions to me seems a bit extraneous. I've put forward an idea that possibly an indent would more suitably replace the bullet point, since it applies white space and general design alignment principles, meaning that it blends in more and is more visibly akin to the previous item for people who are just skimming over the infobox. ItsPugle (talk) 11:58, 31 March 2020 (UTC)

It's probably worth looking at Infobox settlement for guidance, since this template should be as similar to that one as possible in order to provide a consistent reading experience to our readers (who vastly outnumber editors, let alone editors who are dedicated enough to care about infobox formatting). Infobox settlement uses a bunch of bullets for various sub-entries. – Jonesey95 (talk) 14:36, 31 March 2020 (UTC)

Region coding too ?

Each town has been individually sub-cat-coded into it's regional area sub-cat, Towns in the sub region, but the infobox adds in Towns by state by default, so now every place is doubled, sort of. If you can require a state can you get an area/region too and auto add it into the sub region ? it would need every one to be edited, and info added, and sub-cats removed and all Dave Rave (talk) 09:21, 16 June 2020 (UTC)

@Kerry Raymond: all towns are here from the template, all towns are manually coded to the sub cats Dave Rave (talk) 23:44, 19 June 2020 (UTC)

How do you embed other Infoboxes?

I want to embed other infoboxes. How is that done? I cannot see support for module, child nor embed. Betterkeks (talk) 01:36, 12 July 2020 (UTC)

Logo placement

Can we move {{{logo}}} on top of {{{image}}}? ...or below everything? Having it between rows |dataX= and |dataY= is aesthetically bad and is very disorienting to notice or to read maybe, e.g. City of Mandurah. – McVahl (talk) 09:40, 23 July 2020 (UTC)

What broke it?

@AussieLegend: The infobox is appearing on the left without a box. It is doing this on both my PC and my iPad, so I don't think it is anything to do with my settings (not that I have changed them recently). This seems to have happened this afternoon. There have not been any recent edits to the Template according to the history so guessing there is a problem in one of the templates it uses. Kerry (talk) 07:15, 5 September 2020 (UTC)

It looks like someone fixed whatever they broke. Everything is back to normal. Kerry (talk) 07:36, 5 September 2020 (UTC)

It was site-wide according to Wikipedia:Village pump (technical)#Sitewide bug. Kerry (talk) 07:38, 5 September 2020 (UTC)

Template-protected edit request 2020-09-21

Despite edit some days ago [8] still appears in Category:City infobox templates where it is the only template. It isn't restricted to cities and is fine upward in Category:Australia place infobox templates which is inside Category:Place infobox templates by country. TerraCyprus (talk) 22:17, 21 September 2020 (UTC)

To editor TerraCyprus:  done. When this type of edit is made, it sometimes takes several days for the template page to clear the queue. So in the future when you want a more immediate change, try requesting a null edit on the template page (not the /doc page) and see if that clears the category of the template. P.I. Ellsworth  ed. ^put'r there 01:35, 22 September 2020 (UTC)

To editor Paine Ellsworth: thank you for the pointer to Null edit. Good to know a Null edit may help. TerraCyprus (talk) 15:42, 22 September 2020 (UTC)