Talk:Import/Catalogue/ibpt belgium antennas
Issue with telecom=antenna usage
Dear promoters and maintainers of this import. Thank you to get involved on telecom infrastructure importing in OSM, I've just discovered this page and your great results.
I spent significant time earlier, back in 2018 and 2019, with @ZZ29: and others to cleanup telecom=* tagging and antenna wasn't documented yet. A huge amount of features remains to be refined, radio communications is part of them with many equivalent tags in man_made, communication...
As I understand what was done during this import, it regards only sites/places like towers, masts, building roofs and not actual antennas, is that correct?
If yes, it's not so consistent with existing practices, with man_made=tower for instance. tower:type=communication is still used at 200k occurrences and changing it is possible but should be discussed at world scale.
It's a big concern for me, I'm sorry to not having seeing this work before can I get in touch with you please? Fanfouer (talk) 23:08, 29 November 2020 (UTC)
- Pinging the import leader @Vucod:.
The import indeed considers the sites, not individual antennas. The tagging information on the import page is outdated. The current information is at Wikiproject Belgium/Infrastructure/BIPT antennas. In short: All get communication:mobile_phone=yes and the refs. A dedicated mast gets man_made=mast tower:type=communication. A separate station such as on a rooftop gets telecom=antenna. A station mounted on a node such as power=tower doesn't get extra tags.
Man_mast=mast isn't used for stations on rooftops unless there's actually a real mast on the rooftop (such as 6497362560
6497362560, but it's not for mobile phones).
Does this address your concern? Would you tag the rooftop station differently? Best —M!dgard [ talk ] 11:36, 7 December 2020 (UTC)
- Sorry, we had a long discussion in private. I didn't report it here yet. During the import, we assumed that a site will hold at least an antenna. Therefore, we included the telecom=antenna tag until manual review, where it could eventually be replaced by a tag for the supporting structure. @Fanfouer: pointed out that this assumption could be not accurate in certain cases. For example, a site for antenna in a tunnel might be located far from the actual position of an antennas. I agreed on this. The manual review, which is ongoing, should solve this. @Fanfouer: also highlighted the need to clearly distinguish the sites (contain the equipment), the supports and the antennas. In this import, we had nodes for either an antenna+site or a mast+site. Again, upon manual review, this scheme could be improved by separating the site from the antenna or mast where it is needed. Also, more info on the antennas could be added like the quantity, the locations,... But with the info we had from the BIPT, it was the best way we found to import the info in OSM. @Fanfouer: please, correct if I'm not reporting correctly our discussion or if you have other concerns. I'm again open to suggestions or improvements. --Vucod (talk) 10:32, 9 December 2020 (UTC)
- Hi guys and thank you for this discussion. My concern sounds to be well understood here. I'd be in favour of a remains-to-be-proposed telecom=radio_site to state a telecommunication service is provided with one or several antennas at a given place (node, area, relation). telecom=antenna should be kept to tag individual antennas devices. How do you feel about moving telecom=antenna to telecom=radio_site for IBPT sites remaining to be reviewed? Once reviewed, you could tag a mast man_made=mast + communication:mobile_phone=yes and use telecom=radio_site on surrounding fence. It's hard work and I respect what you've done so far here.
- For instance here, there are at least 14 antennas and a single mast. In many situations, tagging each antenna around a single support will be hard work on OSM but could be easily achieved when individual antennas are installed at building corners for instance (or at tunnels outlets) Fanfouer (talk) 22:54, 9 December 2020 (UTC)
- Sorry, we had a long discussion in private. I didn't report it here yet. During the import, we assumed that a site will hold at least an antenna. Therefore, we included the telecom=antenna tag until manual review, where it could eventually be replaced by a tag for the supporting structure. @Fanfouer: pointed out that this assumption could be not accurate in certain cases. For example, a site for antenna in a tunnel might be located far from the actual position of an antennas. I agreed on this. The manual review, which is ongoing, should solve this. @Fanfouer: also highlighted the need to clearly distinguish the sites (contain the equipment), the supports and the antennas. In this import, we had nodes for either an antenna+site or a mast+site. Again, upon manual review, this scheme could be improved by separating the site from the antenna or mast where it is needed. Also, more info on the antennas could be added like the quantity, the locations,... But with the info we had from the BIPT, it was the best way we found to import the info in OSM. @Fanfouer: please, correct if I'm not reporting correctly our discussion or if you have other concerns. I'm again open to suggestions or improvements. --Vucod (talk) 10:32, 9 December 2020 (UTC)
