Template talk:Germany district OSM map

Bot run

@DannyS712: Hi, the bot run was approved, the template can be deployed now.--eh bien mon prince (talk) 06:48, 11 April 2019 (UTC)[reply]

@Underlying lk: I already did a bot run... --DannyS712 (talk) 06:49, 11 April 2019 (UTC)[reply]

@DannyS712: That's odd, this tool shows only 80 transclusions or so, and most pages about German districts such as Oberbergischer Kreis still don't have this template.--eh bien mon prince (talk) 06:57, 11 April 2019 (UTC)[reply]

Then I don't know, sorry - if you can get me another list, I can run it again. What I need is: the page to tag; and the QXXXX id to tag it with. I don't have the time or knowledge to make the list, but if you make one I should be able to try again. --DannyS712 (talk) 07:04, 11 April 2019 (UTC)[reply]

Here's the list.--eh bien mon prince (talk) 07:11, 11 April 2019 (UTC)[reply]

Thats what you got last time, and it didn't really work. Some of the pages are duplicates, and a bunch don't exist on the english wikipedia. If you can't do it all online, you can email me a table with the data, but I need (ideally just) the page names on enwiki and the Q ids you want them tagged with. Sorry, --DannyS712 (talk) 07:17, 11 April 2019 (UTC)[reply]

@DannyS712: Try this.--eh bien mon prince (talk) 07:56, 11 April 2019 (UTC)[reply]

Thanks, I'll try to use that next time I have the chance to work on that task. --DannyS712 (talk) 08:09, 11 April 2019 (UTC)[reply]

Okay, I think I can do another run. Does User:DannyS712 test/infobox5.json look right? --DannyS712 (talk) 04:28, 14 April 2019 (UTC)[reply]

@DannyS712: yes, looks fine.--eh bien mon prince (talk) 08:06, 14 April 2019 (UTC)[reply]

@Underlying lk: I probably won't have time tomorrow - please ping me if I forget to do it in a few days --DannyS712 (talk) 08:08, 14 April 2019 (UTC)[reply]

@Underlying lk: I ran a semi-trial, and made 13 edits ([1]). Can you take a look? I don't see anything wrong, and if you don't either, I'll do a full run. Thanks, --DannyS712 (talk) 23:27, 15 April 2019 (UTC)[reply]

@DannyS712: Mostly fine, but it's probably better to use this query, to avoid abolished districts: link.--eh bien mon prince (talk) 23:44, 17 April 2019 (UTC)[reply]

@Underlying lk: I understand your point, but the overhead of converting the query to the format for the bot is pretty high it its just going to result in a few pages being removed from the list (the bot will just skip any that are deleted) - is there any other reason not to use the current dataset? --DannyS712 (talk) 23:48, 17 April 2019 (UTC)[reply]

That's the only issue I found in the test edits.--eh bien mon prince (talk) 23:51, 17 April 2019 (UTC)[reply]

Are those edits wrong? Wouldn't it make sense for even former districts to have maps added? --DannyS712 (talk) 23:56, 17 April 2019 (UTC)[reply]

OSM only shows current districts, so the template will display a map, but the subject of the article won't appear in it.--eh bien mon prince (talk) 12:01, 18 April 2019 (UTC)[reply]

@Underlying lk: oh, okay. I'll work on the new data set soon --DannyS712 (talk) 16:35, 18 April 2019 (UTC)[reply]

@Underlying lk: I did another round - can you check [2]? Also, I'm seeing that a major limitation is the regex - currently it only works when there is already a map parameter supplied to the template for it to replace. If there is no (even black) |map=, no changes are made. I'll look into a regex that fixes that, but for now can you take a look? --DannyS712 (talk) 01:38, 19 April 2019 (UTC)[reply]

All edits in the last batch are fine.--eh bien mon prince (talk) 08:51, 19 April 2019 (UTC)[reply]

@Underlying lk:

Done the rest of that batch. --DannyS712 (talk) 01:02, 21 April 2019 (UTC)[reply]

Great, looking forward to the rest.--eh bien mon prince (talk) 08:17, 21 April 2019 (UTC)[reply]

@DannyS712: many districts included in the query (such as Limburg-Weilburg) still don't have a map - can you take care of this?--eh bien mon prince (talk) 11:35, 6 May 2019 (UTC)[reply]

@Underlying lk: its the regex - I still haven't figured out a good way of adding a map when there isn't one, short of putting it on the first line of the infobox - would that be okay? --DannyS712 (talk) 13:03, 6 May 2019 (UTC)[reply]

That's less than ideal, but still better than leaving so many articles with no maps.--eh bien mon prince (talk) 23:28, 8 May 2019 (UTC)[reply]

@DannyS712: Let's fix it any possible way - better than leaving it like this.--eh bien mon prince (talk) 15:24, 19 May 2019 (UTC)[reply]

@Underlying lk: okay - I'll finish the run soon --DannyS712 (talk) 18:48, 19 May 2019 (UTC)[reply]

@Underlying lk: Sorry it took so long - how does User:DannyS712 test/infobox7.json look for the additions? I'll remove anything that has the template already, or that doesn't have the infobox --DannyS712 (talk) 03:05, 17 June 2019 (UTC)[reply]

@Underlying lk: Hello? I'm ready for infobox7 if it looks good --DannyS712 (talk) 23:23, 21 June 2019 (UTC)[reply]

@DannyS712: It doesn't - this template is for districts, I see many cities like Frankfurt in the bot run list. Try this query instead.--eh bien mon prince (talk) 16:34, 29 June 2019 (UTC)[reply]

@Underlying lk: I used the updated query - again at User:DannyS712 test/infobox7.json. Look good? --DannyS712 (talk) 16:41, 29 June 2019 (UTC)[reply]

@DannyS712: Yes, this should do it.--eh bien mon prince (talk) 16:42, 29 June 2019 (UTC)[reply]

@Underlying lk: Okay, I made 7 test edits (see [3]) and think they look good. If you can take a look, I'll do the run. But, just a warning that around 75% of the pages in infobox7.json are being skipped because they either already have the map template or don't have the right infobox (lots use {{Infobox German location}}, which isn't in the scope of this task). Once you take a look and I do the run, I'll file the Russia brfa. --DannyS712 (talk) 18:02, 29 June 2019 (UTC)[reply]

@DannyS712: All edits look okay.--eh bien mon prince (talk) 18:17, 29 June 2019 (UTC)[reply]

@Underlying lk:

Done. If you don't see anything wrong, I'll file the next BRFA. Sorry this took so long of back and forth --DannyS712 (talk) 18:49, 29 June 2019 (UTC)[reply]

Some districts are not working

Hi, I saw that the map template does not work correctly for the districts of Wetteraukreis and Main-Kinzig-Kreis. In both images, neither of the two districts shows up. Unfortunately, this makes it look like the districts are what is marked green on those maps. So an unaware user would think that the Wetteraukreis is the whole of Hesse except for the Wetteraukreis and Main-Kinzig-Kreis.

Interestingly, there are other maps, e.g. for the Vogelsbergkreis, where both districts show up (but they are obviously not highlighted), and there are other maps, e.g. for Schwalm-Eder-Kreis, where they also do not show up. Spike (talk) 08:26, 29 July 2019 (UTC)[reply]

After looking through the pages of all Hessian districts, it looks like there is an additional third problematic district which only appears in a minority of all district diagrams: In addition to the already mentioned Wetteraukreis and Main-Kinzig-Kreis, there is also the Vogelsbergkreis. And it's always the case that those three districts are either all three missing from the diagram or all three visible. Interestingly, the Vogelsbergkreis is visible on its own page, but not visible on some other pages. Wetteraukreis and Main-Kinzig-Kreis have the additional problem that they are not visible on their own pages, leading to the problem described above. Spike (talk) 16:24, 29 July 2019 (UTC)[reply]

This is strange, now Vogelsbergkreis has the same problems as the other two: it is also not visible on its own page. I could have sworn that was different two days ago. Also, pinging DannyS712 and Underlying lk, because you seem to have done the development of this template. Spike (talk) 08:31, 31 July 2019 (UTC)[reply]

There are more districts with similar problems:

Mecklenburg-Vorpommern: Vorpommern-Greifswald and Rostock are not shown
Hesse: Main-Kinzig-Kreis, Wetteraukreis and Vogelsbergkreis are not shown
Saxony: Freital is shown as a separate district, but it is part of Sächsische Schweiz-Osterzgebirge
North Rhine-Westphalia: Wattenscheid (part of Bochum) is shown darker than the rest of Bochum, and part of Velbert is shown darker than the rest of Mettmann (district)
Bavaria: Landsberg am Lech is shown as a separate district, but it is part of Landsberg (district). Same for Freising and Bad Reichenhall. The town Deggendorf is shown darker than the rest of Deggendorf (district)

Could someone who is familiar with this kind of OSM templates (Underlying lk, DannyS712, Begoon?) check what's the matter? Markussep ^Talk 18:15, 6 September 2019 (UTC)[reply]

How the template works: it is basically a Wikidata query, much like this one, that selects all entities that are an instance/subclass of independent city of Germany (Q22865) or district of Germany (Q106658) - with some exceptions, such as when end time (P582) or dissolved, abolished or demolished date (P576) are set on the page. The parent state is either provided manually or found through the property located in the administrative territorial entity (P131).

All of this only works if the entity is connected to an OSM relation: Rostock (Q2861) is selected by the query above, but the OSM relation Kreisfreie Stadt Rostock is connected to Rostock (Q32265531) (as visible here), so it makes sense that it would not be displayed. Just merging the two WD entities and pointing OSM to the right one would be enough to make it appear again.

Vorpommern-Greifswald is a different case - Vorpommern-Greifswald District (Q2917) is correctly selected by the query, the OSM relation points to the right entity, but it still doesn't show up. Overpass Turbo displays it, but gives out a warning about its file size. So it might not be shown on Wikipedia due to that. I'll have to look into the issue more to see if it's fixable, or if it's advisable to go back to non-interactive maps.--eh bien mon prince (talk) 07:52, 14 September 2019 (UTC)[reply]

Thanks for your reply. I checked what could be the problem with the three Hessian districts, but they all three show up in the Wikidata query, and their maps show up in the overpass tool as well (no warning like the one for Vorpommern-Greifswald). Freital shows up in the query for Saxony, while it shouldn't. It used to be an urban district (Q22865), but it isn't anymore (end date P582 given: 1946). Does the query miss the P582 property for urban districts? Markussep ^Talk 18:57, 14 September 2019 (UTC)[reply]

An update:

Mecklenburg-Vorpommern: Vorpommern-Greifswald is shown now, Rostock isn't (OSM link needs to be corrected)
Hesse: Main-Kinzig-Kreis, Wetteraukreis and Vogelsbergkreis are shown now Y
Saxony: Freital is still incorrectly shown as a separate district (and still pops up in the wikidata query)
North Rhine-Westphalia: Wattenscheid (part of Bochum, formerly an urban district) is still shown darker than the rest of Bochum (and still pops up in the wikidata query), the problem in Mettmann (district) (Velbert-Neviges) is fixed now
Bavaria: the former urban districts Landsberg am Lech, Freising, Bad Reichenhall and Deggendorf are still incorrectly shown (and still pop up in the wikidata query) Markussep ^Talk 09:40, 7 January 2020 (UTC)[reply]

More broken districts

I have found some more districts which do not seem to be working: Mansfeld-Südharz and Salzlandkreis in Saxony-Anhalt, Göttingen (district) in Lower Saxony and numerous districts in North Rhine-Westphalia (from what I can tell Rhein-Kreis Neuss, Mönchengladbach, Essen, Viersen, Krefeld, Ennepe-Ruhr-Kreis and Dortmund). None of them show up on any of the district maps for their state but there doesn't seem to be anything obviously broken about the entries. I checked the Saxony-Anhalt ones and Göttingen in Overpass Turbo and seem to show up correctly in the query. Also on the Lower Saxony map both the cities of Hanover and Göttingen show up (possibly due to the "Göttingen Law" and "Law on the region of Hanover" mentioned in Lower Saxony#Administration) while on the North Rhine-Westphalia one the city of Aachen and the afore mentioned Wattenscheid show up in addition to their districts. Alphathon /'æɫ.fə.θɒn/ ^(talk) 05:49, 2 November 2021 (UTC)[reply]

It would seem Plön, Rendsburg-Eckernförde, Segeberg and Steinburg in Schleswig-Holstein are also not working. Alphathon /'æɫ.fə.θɒn/ ^(talk) 17:12, 2 November 2021 (UTC)[reply]