Proposal talk:Cable Landing Station

From OpenStreetMap Wiki
Jump to navigation Jump to search

Replace man_made=* with a more appropriate key

Despite I agree that a cable landing station is a man made artifact, we should use a more appropriate key to get a better focus on the matter.
Too many different things are built by Human and they are so numerous that the man_made=* doesn't bring the information it is supposed to.
As such stations are definitely a telecommunication infrastructure, I propose to change for telecom=*. telecom=cable_landing_station may be good but a bit long. telecom=cable_station may be even better and suitable for terrestrial cables too. Let's discuss about that Fanfouer (talk) 19:13, 10 March 2026 (UTC)

I'm not necessarily in favor of man_made=cable_landing_station, but when I posted on the community.openstreetmap.org, this was the first suggestion to be repeated. telecom=cable_landing_station was also suggested, then the idea of matching the pattern used by other linear networks was put forth (e.g., telecom=facility + facility=cable_landing_station). I personally like the linear network pattern the best as it mirrors two other major infra segments: power=* and man_made=pipeline. The current state of telecom=* is interesting: it mixes function tags to be paired with other objects (e.g., exchange, connection_point, data_center) with stand-alone objects (e.g., distribution_point, service_device) and linear objects (line which overlaps with communication=line). I don't disagree telecom=cable_landing_station would work (and requires the least coordination with the OSM telecom=* community), but I wasn't sure if it was best to continue this mixed pattern or to suggest something like the facility=* idea or telecom:function=cable_landing_station to see if there was interest in changing the path telecom=* has been going down for the past 10 years or so. -- Trailrunner13 (talk) 19:48, 10 March 2026 (UTC)
Thank you for elaborating here. I wonder from where this idea with facility=* comes? We don't have such for power=* nor pipelines. It hasn't appeared in Proposal:Utility_facilities in particular.
It's fine to me to mix facilities and devices in the same key (power=substation vs power=transformer) since they are well defined and properly documented as such.
Telecom lines hasn't been reviewed yet, which explains why telecom=line still overlaps communication=line. the last should be removed.
telecom=facility won't avoid such mixing anyway. Fanfouer (talk) 22:02, 10 March 2026 (UTC)
I understand what you are saying. Let me try to explain... the explanation wanders a little, but hang in there:
- man_made=pipeline is a "second level tag" used to mark a linear object plus pipeline=* for related physical objects like pipeline=substation which often have an associated function. There is also has a separate tagging scheme under usage=* for pipeline functions. Seems pretty complicated. Unrelated question: why didn't people simply choose pipeline=line instead of putting this into man_made=*? >:-|
- power=* is a "top level tag" with what appears to be only stand-alone objects. Like pipeline=*, there's power=substation which can have functions (e.g., transmission vs. distribution). Interestingly, power=* also uses physical differences to describe function: power=line means "transmission" and power=minor_line means "distribution".
- telecom=* is a "top level tag" and mixes stand-alone objects and functions. When I say "stand-alone", that means it doesn't need to be combined with another tag (e.g., telecom=distribution_point ) while a "function" modifies something else (e.g., building=service + telecom=exchange).
- Some in the community suggested since there was: A) pipeline=substation + substation=compression or substation=distribution and B) power=substation + substation=transmission or substation=distribution, maybe telecom=* could also follow that pattern.
- If we did that, there could be telecom=facility + facility=cable_landing_station or facility=exchange or facility=data_center. "facility" is probably a bad word... "function" is probably better because that's what we are really trying to describe. Or use a single tag notation: telecom:function=function_name.
- This got me thinking further about separating object and function. If there was telecom:function=*, why not also have telecom:structure=*?
- If we did that, we'd have telecom:function=cable_landing_station to go with telecom:function=exchange and telecom:function=data_center. I'd then like to add telecom:function=amplifer and a few other optical network functions.
- We'd also have telecom:structure=distribution_point and telecom:structure=service_device.
- telecom=connection_point appears to be the same thing as man_made=street_cabinet + utility=telecom, I'm not sure how to think about that. Or maybe telecom=connection_point was meant to be a function? Or is everything in telecom=* meant to be a function?
- This is not in the Proposal, but telecom=* could also leverage usage=* like pipeline, and have usage=cable_landing_station, usage=exchange and usage=data_center.
-- Trailrunner13 (talk) 03:07, 11 March 2026 (UTC)
Thanks again for completing your answer, we will surely find a suitable solution for the proposal. OpenStreetMap needs such discussions and the tagging model is continuously evolving. We have got a pretty long history in that business and some past choices should be considered with this in mind.
There is no top level or secondary tags. What you are seeing here is containers (man_made=street_cabinet, building=service + utility=*) and functional roles (power=*, telecom=*...) both for facilities and devices. Sometimes we only got the container and sometimes we miss a suitable definition. That's why you've got the impression there may be standalone objets. Particularly for telecom=distribution_point, it could be contained on man_made=utility_pole or any suitable tag to be defined for stalls or pedestals.
man_made=pipeline is a quite old tag, introduced in 2007 and as you guess, it should be refined with pipeline=line or something to give up the man_made=*. Such work hasn't been done yet as we are all volunteers and there are more urgent tasks to be done now. So is power=pole, which implies man_made=utility_pole + utility=power for consistency sake without ability to replace it in practice.
That's why I reasonably think what we need here is building=service + utility=telecom + telecom=cable_station and it has great chances to be approved as it's a very simple solution to be implemented in the existing landscape. Fanfouer (talk) 11:18, 11 March 2026 (UTC)
After marinating on this for a day, I agree the most practical solution is to adopt telecom=cable_landing_station, then have the broader discussion around whether now is the time to reform telecom=* (I will likely prepare a separate Proposal for this). I can update the man_made=cable_landing_station page to reflect this change.
As for the shorter, but less specific cable_station, this is not a term I've heard during a long telecoms career. I also checked my copy of Newtons's Telecom Dictionary just to be sure. Because a submarine cable is unique (it contains both optical fiber and a copper electrical conductor), terrestrial networks don't have sites quite like the CLS which combines SLTE (specialty long reach coherent DWDM, muxes and add / drop equipment) and PFE (high voltage DC electrical equipment to power optical amplifiers laying on the seabed).
In long haul terrestrial networks, there are equivalent optical networking functions, but nothing like the power function, and the terminology used is different: "in-line amplifier" sites ("ILA" or just "amplifier") are most common, but there are also sites with 3R functionality and add / drop functionality with the latter usually ending up in a "point of presence" (PoP) site.
As it currently stands, telecom=*'s current language is tilted towards copper networks... see the Proposal's Problem Statement. While the most expedient path is to simply add telecom=cable_landing_station, this will introduce an optical networking term to telecom=* which leads to eventually adding telecom=amplifier and telecom=point_of_presence. But then what happens when people want to add the unique language used in the HFC (Cable TV) industry (e.g., telecom=head_end, telecom=cable_modem_termination_system), metro fiber / FTTP industry (e.g., telecom=optical_line_terminal) and so on?
All of this doesn't need to be determined now, but I wanted to share another aspect for why I was suggesting ideas like telecom:function=* and telecom:structure=*.
Any other suggestions before I update the man_made=cable_landing_station page? Then I guess I change update Proposal Status and announce the Proposal? I've not done this before, so any process advice is welcome.
-- Trailrunner13 (talk) 13:48, 12 March 2026 (UTC)
I'm fine with telecom=cable_landing_station for this round. You may find the whole process to follow for a proposal to be reviewed on the Proposal_process. The page man_made=cable_landing_station should be deleted for now and reviewed tagging during the vote will be documented on according pages after post-vote cleanup. There is no need to create new wiki pages during the proposal's review.
CMTS, DSLAM and OLTs are currently addressed with telecom=service_device + telecom:medium=*.
I'm also fine with thinking about adding new values for telecom=*, particularly for missing features like amplification rooms and point of presence. But this should be done on separate proposals to keep consistency in discussions. Feel free to ping me when ready for this Fanfouer (talk) 16:32, 12 March 2026 (UTC)
Our discussion made me draft this page about tagging infrastructures we were missing for years. Hope it helps Fanfouer (talk) 23:26, 12 March 2026 (UTC)
Resolved: shifted from man_made=* to telecom=* with other structural improvement ideas deferred for now -- Trailrunner13 (talk) 14:32, 17 March 2026 (UTC)

Next topic?