@felipeeugenio

"Don´t invoke local communities."

Why?

"Do the Chilean mappers have a say in this?"

Of course! Please feel free to comment, and as someone who has been watching the region for a long time, I'm taking your point of view into consideration, even if I disagree. For me, your opinion counts, adding to those on the tagging mailing list. Together, they all currently lean towards your version. If within a few days the consensus continues to indicate that, I will update the wiki and downgrade highway classification in all of Antarctica uniformly to reflect the consensus.

"Did you discuss with the Chinese and Uruguayans?"

There are very few mappers in Antarctica, so I haven't found any yet other than casual mappers uninterested in highway classification. In fact, over two years, you are the first Chilean mappers I have come across in Antarctica. I've seen mainly Europeans from various countries. The initial mappers of the Fildes Peninsula no longer appear to participate in OSM.

"this is another "I own the place""

Quite the opposite. You can check the history of highway classifications in Antarctica in multiple places, not just the Fildes Peninsula. When I started, there were big discrepancies between these places: what you just changed to tertiary was similar to another road somewhere else that was mapped as service, somewhere else as track, somewhere else as residential, somewhere else as highway=road, etc. I welcome suggestions on how to improve the result. I don't think different parts of Antarctica should follow different rules, but the rules are up for debate, and as far as I know, this is the first time the are being debated specifically for Antarctica, which is a unique place in the world.

@felipeeugenio

"How about a complete survey?"

I actually did this a long time ago, but here are some other examples of similar regional configurations:
- Norway: Gamvik, Fredvang
- Faroe Islands: Tjørnuvík, Kunoy
- Iceland: Flateyri, Tálknafjörður, Hjalteyri
- Canada: Whatì, Tulita
- United States (Alaska): Eagle, Teller
- Russia: Varzuga, Yamburg, Chasovenskaya

@felipeeugenio

"which included several names that you removed after some months."

I removed information based on lack of verifiability as I have mentioned in my changeset comments. I have always been very lenient in Antartica as the data sources are difficult to come by, but I have seen many really bad edits all over it, including mappers using it as a scratchpad and leaving junk behind. So, in order to maintain data reliability, if you can provide a source, I will readd those names or features after verifying them and also add the source to the wiki: osm.wiki/Antarctica/Data_Sources

"We had to correct an unclassified road into Mount Vinson from you."

I don't fully agree with the correction, but none of us have access to data that demonstrates which mapping is more correct, we only have the Strava running heatmaps, where many tracklogs match motorised routes elsewhere in Antarctica, so I think it is not worth discussing this case without better data.

"And even the Svalbard example that you are citing, they are using a secondary (wrongly) to join two sectors of the same urban settlement. But as soon as the roads leave the Village they become unclassified roads."

Using highway=secondary to connect place=suburb is generally correct as far as I know if we don't assume any country's specific highway classification scheme. Ways that continue into sparser areas past a settlement's core have lower classification in Svalbard as they did in the Fildes Peninsula before these recent edits, so the pattern was essentially the same.

"Then again, please join the Chilean community."

Thank you, I will look into that. But as Antarctica is international space under the Antarctic Treaty, I think any discussion on it should involve communities from multiple countries, be it in Fildes, McMurdo or other areas. Since each local community has its own ideas about highway classification, disagreements are inevitable and it's best to find consensus rather than impose any country's specific system. While the Chilean settlement in the Fildes Peninsula is the largest and oldest, the peninsula is also home to year-round Russian, Chinese and Uruguayan stations.

"The eastern part, which starts as primary (wrongly), then leaves town as a tertiary just to head to a radar and the OREO Doomsday Vault. That is wrong too."

I think it would generally make sense for this route to be highway=tertiary up to the entrance of the Camp Beretz settlement (currently not mapped in OSM) and also all the way up to the Gruve 7 coal mine which likely attracts heavy traffic.

@jptolosa

Why would tertiary be the highest classification that can be used in the Fildes Peninsula?

"A reasonable solution can be using tertiary to join aerodrome with Villa Las Estrellas and using unclassified to connect Villa Las Estrellas with the bases that depending of the village."

I disagree with this based on the text in the wiki about highway=* assumptions, particularly this part: "In a region with poor infrastructure, a road of highest importance, forming the main road network there, should be highway=trunk, regardless of being a high-quality wide asphalt road or a low-quality narrow track worse than highway=service in other regions." These roads constitute the region's main network and are vital for its functioning, so their classification should not be so low.

For comparison, the Norwegian Arctic island of Svalbard has about 2,600 inhabitants, one airport and one larger town (Longyearbyen) of 2,100 inhabitants. A highway=primary road connects the town to the airport. The winter population in the Fildes Peninsula is about 115 people based (from COMNAP's 2017 Arctic Station Catalogue, referenced in Wikipedia), or about 1 order of magnitude below that of Svalbard, so it makes sense to use highway=secondary for an analogous connection. A highway=secondary way connects Longyearbyen it with the small Nybyen suburb which is similar in size (number of buildings) to Artigas Base and Great Wall Station, so again it makes sense to use highway=secondary for these analogous connections.

"The nearby bases have little or zero permanent population."

If permanent population becomes a criterion in Antarctica, then there will be no Antarctic roads with a class higher than highway=unclassified. McMurdo Station has a wintering population of 153 people and Amundsen-Scott Station has a wintering population of 49 and neither have permanent residents, yet there is little doubt that the South Pole Traverse, the most important (ice) road in Antarctica, should be displayed very prominently on the map.

Many local communities have adapted their thresholds for classifying sparsely populated places. Several local communities employ highway classification systems based on place hierarchy. And some even allow these population thresholds to vary across the country's territory, in order to achieve a more sensible and useful highway classification. Advocating for applying some sort of worldwide default population threshold to the sparse regions of Antarctica would go against what several local communities have done in the sparse regions of their own countries.

*the previous classification

These buildings are part of and nearest to the Russian Bellingshausen station. To follow the informal convention used in the region for years, their main name should be in Russian.

The previous was based on the relative importance given the local context. See highway=*#Assumptions

Why?

Muitas etiquetas no OSM surgiram espontaneamente. O que definiu se elas ficam foi o status quo mesmo: muitos mapeadores adotaram, então, foi considerada aprovada. Algumas inclusive tiveram oposição tardia, conforme outras pessoas foram entendendo os problemas gerados.

Às vezes acontece que o mapeador inventa coisas que não podem ser adequadamente consumidas nas aplicações. Então há um certo atrito entre o mapeador e o desenvolvedor. Outras vezes o mapeador inventa coisas que outros mapeadores não estão dispostos a manter, geralmente porque deixa o mapa mais confuso ou difícil de editar. Esse é talvez o caso dos elementos subterrâneos: eles são difíceis de verificar a não ser que alguém de fato entre debaixo da terra. Mesmo assim, alguns elementos subterrâneos são permitidos para comporem "redes" com elementos de superfície, como tubulações de água conectadas a rios, e os metrôs. Esses casos geralmente foram introduzidos por importações de dados, até porque é difícil alguém verificar (impossível entrar nas tubulações, e nos metrôs não há sinal de GPS). Importações costumam ter bastante discussão.

Acho interessante o mapa ter essa informação das minas considerando os riscos bem conhecidos dessa atividade para a população local. Talvez as tags usadas nesse caso não sejam as ideais. Sugiro trazer esse assunto no fórum propondo uma forma de mapear, talvez propondo etiquetas próprias (que poderiam ser selecionadas no Overpass Turbo) ou futuramente mostradas em alguma renderização específica, como se pretende com o OpenHazardMap (osm.wiki/OpenHazardMap). Só aviso que o OpenHazardMap é um projeto em curso há anos e ainda não tem visualização própria (há várias etiquetas no OSM nessa situação).

Caindo de pára-quedas, só queria deixar uma sugestão. Em geral, no OSM nós mapeamos o que existe na superfície, em alguns casos há previsão para mapear alguns elementos subterrâneos. No caso de uma mina subterrânea, o que está na superfície são as entradas da mina. Então, acho que existe uma representação possível pras minas de sal-gema (man_made=adit ou man_made=mineshaft, qualquer uma dessas duas junto com resource=salt), o que talvez esteja faltando é a informação sobre a localização exata dessas entradas, o que talvez possa ser obtido com alguma investigação.

Are you sure that Icestock Plaza refers to the building? Maybe it refers to the area between buildings where Icestock takes place, currently mapped as a node here: node/10807908896

If it does not refer to the building, then please remove the name from the building and update the node as needed.

For clarity, I meant "designation" instead of toponym and "toponym" instead of name.

Fora o ponto -25.510405, -54.5858801 na rua Fagundes Varela, o restante da geometria está visível da rua? Se sim, eu gostaria de pedir então fotos do local se você as tiver. Faz sentido que seja mais a oeste do que o que consta no Mapa do Exército mesmo, mas é uma informação difícil de verificar.

O arroio subterrâneo não é visível na imagem de satélite do Bing. Como você sabe que ele não passa na avenida esim debaixo dessas outras quadras?

Limite de velocidade de 10 km/h?

Além disso, ao fazer correções na classificação, não deixe lacunas na malha. Promover os trechos intermediários sem promover junto as travessias urbanas (por qualquer que seja a razão) deixa a malha fragmentada e torna o mapa ilegível e pouco profissional (compare com os mapas comerciais). Procure pensar no sistema viário intermunicipal como rotas ligando nodos viários.

Olá Cleverson. Sua reclassificação foi desfeita por outro mapeador por não seguir a regra de classificação adotada por consenso tanto pela comunidade do RS quanto pela brasileira. As estradas que você marcou como importantes não são rotas principais entre cidades com mais de 200 mil habitantes, mas são entre cidades com mais de 20 mil habitantes, e por isso devem ser primárias, não troncais. Peço também que, ao fazer modificações controversas como esta, explique no comentário da alteração o que foi feito e por quê. osm.wiki/Brazil/Classifica%C3%A7%C3%A3o_das_rodovias_do_Brasil

我的意思是“全年科学考察站”。

南极洲的所有小型永久性研究站都被绘制为小村庄（place=hamlet）。

One can map temporary structures that are frequently rebuilt, such as ice and winter roads, and structures that move slowly, such as drifting ice stations. For example, McMurdo's ice roads, runways and airfields are rebuilt every year, each time in slightly different positions but following an official plan. Ideally, they should be resurveyed whenever they are rebuilt. Approximate positions should be acceptable when resurvey is not possible (typically the case in Antarctica), as long as each change in OSM increases accuracy. If it has been a few years since the last aerial images were available and there are new official maps showing that a feature has been relocated, one should consider whether OSM data is better with accurate but old data that no longer represents reality, or with new approximate data that matches it better. I'll have a look at Whoop Whoop, it's currently mapped but the location is probably inaccurate.