One thing that needs consideration is how do we want to deal with the currently existing elements tagged with emergency=ses_station, amenity=emergency_service or emergency_service=technical. In total there are about 1500 of these objects. As of now, there is consensus that these tags should be deprecated. But that just means that we will edit the documentation in the wiki. Untill now we did not discuss manual or (semi-)automated mass edits. I think we should defenatly include a clear statemeant about this topic in the proposal. If we want to propose to do (semi-) automated or mechanical edits, we would need to make sure that we obey the Automated Edits code of conduct.
I already did some analysis of the situation. I think at least about half the objects could be cleaned up quite easily without manually checking them. But before going into to much detail about the âhowâ I would like to first discuss the âifâ.
My opinion is, that we should do some mechanical edits to clean up as many objects as possible while ensuring that no false matches get edited too. For the rest we could maybe make a maproulette chalenge, but I am not sure about this.
What do you think about this? Should all edits be done by a human? Should everything be edited automatically? Should we filter the objects to only edit some of them automatically and some of them manually? What do you think about using maproulette?
My problem is that we are talking about disaster resonse. Disasters are handled by people in an office at the appropriate agency level. That means that we are talking about a bunch regional offices. Mark those building with emergency=disaster_response to indicate that office directs resources during a disaster and your are done.
What wee are trying to do is catalog resources used to actually carry out that response. The information about where and who owns it will become apparent as we tag existing locations with thier capabilities.
Those could be a mountain rescue yeam, a kennel of human or cadaver dogs to find and recove lost campers. We need to identify the resource and decide which emergency situations apply.
So when it comes to boat capsizing off the coast of New England, someone in-charge can figure out who is available to rescue those on board. That person can quickly find all the equipment, services and personnel. They can quickly decide what combination of resources might be required to find and safety return those people to shore. It could be involve the local volunteer boat owners trained in open-ocean rescue, local fire fighters, state police or helicopter squad from the nearby Coast Guard station working in concert to make it a safe and successful recovery.
No, we are not. As of now, we have the following definition in the proposal:
emergency=disaster_response applies to a station of a not-military organisation that has the main objective to help the civil population during and after natural or anthropogenic disasters by working in the affected area but does not have firefighting or medical service as their main competence. The place is used for storing and repairing equipment (hand tools, trucks, boats, safety gear, âŚ), training the members (volunteers or paid ones) and doing administrational tasks. It is the place where the members start a mission after getting alarmed.
So the proposal is to establish emergency=disaster_response for duty stations of disaster-response/civil protection organisations, not for offices.
A problem we already came accross is that âdisaster responseâ, âcivil protectionâ, âcivil defenseâ and so on are hard to translate into different languages because disaster response is handled different in different countrys.
At some point there was the idea to additionally introduce emergency=disaster_office or something similar to map administrative places. But I thinkt that would be something for a seperate proposal.
If I understand you right you are now talking about a tagging scheme for stations fitting the definition. So you propose to establish a tagging scheme that enables stroing detailed data about the capabilatys of a station. I like this idea but I think we should in the first step focus on finding a tagging scheme for stations of disaster response organisations.
Why exclude military resources? Soliders and national guard personnel are often have many skills useful in a disaster. The Coast Guard probably spends more time rescuing people than maritime law enforcement.
Most of those words are overly broad concepts. Ironicly many of those names were created during the world wars. They reduced the strain on fighting militaries by training civilian to do more mundane tasks on the homefront.
That why I suggested starting with categorizeing tjings like equipment and the disasters they are designed to âfightâ. Youâre more likely discover more meaning generic terms that everyone can agree on.
I agree with Fizzie. If we included military bases into the proposal, why stop there? If we wanted to tag every facility, that may be helpfull in a disaster, we would need to add emergency=disaster_response to:
Construction companys owning excavators/cranes
Plumbers/electritions who can replace broken pipes and cables
Hardware stores that sell tools
fire fighters, police, ambulance stations
âŚ
If you want to query the map for disaster-response capabilaties of an area it is not as easy as searching for emergency=disaster_response. Depending on what exactly you want to find you need to consider including facilitys like the military or fire stations into your search. That can easily be done by something like this in overpass turbo:
[out:json][timeout:25];
(
nwr["emergency"="disaster_response"]({{bbox}});
nwr["amenity"="fire_station"]({{bbox}});
nwr["landuse"="military"]({{bbox}});
);
out geom;
If I understand you right you would like to introduce a very detailed tagging scheme. So something like
facility_has_big_water_pumps=yes + water_pumps_flowrate=20000l/min
for places with big water pumps
facility_has_big_electric_generators=yes + electric_generators_power=200kVA
for places with big generators
facility_has_boats=yes
for places with boats
Of course these tags are not formulated well. They just describe the general idea I understand from your posts. A detailed tagging scheme that stores all this data about exact capabilatys would maybe be nice to have, but I would realy like to divide this from this proposal to keep it simple. And even if we had the detailed tagging scheme we would still need a main tag to group all of the objects (emergency=disaster_response).
The relevant guideline is duck tagging. Basically, if something looks like disaster response and its purpose is to facilitate disaster response, tag it as disaster response. Other permanent facilities can in many cases be mapped as separate objects with their own tags.
I didnât know what duck tagging is but I like the concept. Thank you for linking to it!
Essentially, you tag according to the âduck testâ - if it quacks like a duck, looks like a duck, and walks like a duck, call it a duck.
The military does not quack like a disaster_response, because helping in disasters is only sometimes their duty. The quack of the military is to be prepared for war. The quack of a disaster-response-station is to send people to disasters to help in the effected area.
How do others see this? Do you agree or does the military quack like disaster_response?
Yes, I think youâve been talking at cross purposes. The proposal is for a tag for sites of organisations whose primary purpose is disaster response. This isnât true for the military. And as @Casey_boy pointed out early in this thread, such organisations may not exist in some countries at all. Where they do, emergency=disaster_response could be the primary tag for their facilities.
What @IanH is suggesting is more like a tag that could be added to all organisations that play a role in disaster response, the same way seamark tags might be added to existing objects that play a role in marine navigation. If that was the goal, it would make sense to include the military, but it probably isnât the goal.
On first glance, a tag with an information like plays_a_role_in_disaster_response=yes would be nice as an easy way to find relevant objects near you in case of an emergency but it would also be very difficult to tell what is relevant. That topic is a completly different concept that would need a lot of thoughts and discussion to work it out. In the proposal there already is a section âFuture plansâ describing the idea of discussing a more detailed tagging scheme after emergency=disaster_response gets approved and adopted.
We keep the current definition of emergency=disaster_response and keep excluding the military because disaster_response is not their main competence.
A discussion of a more detailed tagging scheme that may include something like plays_a_role_in_disaster_response=yes is planed for later as described in Future plans.
Is this an acceptable way? Feel free to up/down vote this post or answer to it.
I would not do anything with the second part at all. I wouldnât like to see could_be_used_for_firefighting=yes tagged on every pond, plays_a_role_in_landslide_response=yes on every shovel or plays_a_role_in_water_rescue=yes on each motorboat.
These things can just be mapped as what they are, and data users can interpret the data freely.
I agree entirely, but having said that, when mapping in bush-fire prone rural areas, I do map backyard swimming pools & similar (with access=private) so that they are visible to rural firefighters, including helicopters, as a potential water source!
I think the consensus right now is that we should keep the current definition for emergency=disaster_response. Applying this tag to everything that might be helpfull in disaster situations would just not be sensfull.
One thing that I meantioned but didnât realy get feedback on is how we want to deal with the current database entrys using emergency=ses_station, amenity=emergency_service or emergency_service=technical. These are the tags that we agreed to deprecate. (How) do we want to change these objects? Edit every one manually or do same mechanical or automated edits?
I did some analysis to help find an answer to this question. First I created a table containing every object using one of the 3 to be deprecated tags using overpass turbo.
overpass query
[out:csv(::user, ::version, ::timestamp,'amenity','emergency_service','emergency','name','operator','brand','addr:city','addr:country','addr:housenumber','addr:postcode','addr:street';',')][timeout:25];
(
nwr["amenity"="emergency_service"]({{bbox}});
nwr["emergency_service"="technical"]({{bbox}});
nwr["emergency"="ses_station"]({{bbox}});
);
out meta;
name=*THW Ortsverband* and operator=Bundesanstalt Technisches Hilfswerk (THW)
688
operator=NSW State Emergency Service
214
operator=Protezione Civile
126
Of the ca. 1500 objects nearly 700 have the THW as the operator=* and âTHW Ortsverbandâ included in the name=*. âOrtsverbandâ means âlocal stationâ, these are the ones that fit emeregncy=disaster_response. They are always called like âTHW Ortsverband Leipzigâ, where âLeipzigâ is the city name where it is located. So I think if we would do the following automated edit, we would not introduse any new errors into the database:
Search for any object tagged with (emergency=ses_station or amenity=emergency_service or emergency_service=technical) and name=THW Ortsverband ...
Remove emergency=ses_station, amenity=emergency_service and emergency_service=technical from the object
Add emergency=disaster_response to the object
The other about 110 objects that have opertor=Bundesanstalt Technisches Hilfswerk (THW) are probably offices, warehouses and so on. So they should not get emergency=disaster_response. I am not sure if I would touch them at all as we do not have anything to replace the current tags with.
What do you think about the idea of this automated edit? Can you think of more automated clean up? Should any automated edit be done? Or should we check every object manually before editing?
Following the guidelines, mechanical edits should be excluded from tagging proposals, so people can vote for the tags without being forced to agree or disagree with a mechanical edit. For mechanical edits we rely on consensus rather than votes.
My suggestion is to discuss the plan for a mechanical edit in the countries where the deprecated tags are most common, and changing the tags based on local community consensus. For countries in which the deprecated tags are uncommon, I would just update them without bothering with lengthy discussions.
Also, a vote result is never permission for large-scale re-tagging of existing objects
It seems that you are right. So basically no matter how we feel about automated or large scale edits, it is not part of the proposal. If I understand it right, as soon as emergency=disaster_response gets approved we can start discussing such large scale edits.
