Hi all, 1. I apologize for closing the proposal during this discussion. It was not due to ignorance. For some reason, Gmail doesn't show all emails from this mailing list. (I Googled for it a couple of times, but couldn't find anything. Does anyone have a clue?) The last email I saw was Warin's answer to Pieren's questions from 13 January. No response appeared in my Gmail, so I went on with the standard procedure and closed the proposal. Today, after reading a seemingly disconnected post from althio, I went to check the tagging list archive and discovered all emails from yesterday and today.
2. Having said this, I would like to draw your attention to the fact that people who currently actively oppose the proposal have not participated in a 4-month discussion, where most of the current concerns were raised and analysed. At the same time, those who participated earlier don't join the current discussion. I could understand if they found it a waste of time and, honestly, I don't understand why you guys were silent for so long. Pieren indeed posted one comment in the discussion page, to which I answered and haven't received any further feedback until now (3 months later). 3. Someone mentioned that other discussions took more than a year. I haven't decided to close the discussion after 4 months. It simply converged and actually someone else proposed to go for voting (thus the group was >1 person, Marc :)). So this discussion once again shows the problems in the current proposal process. 4. To cool things down: Even if the participants of the re-started discussion all vote against the proposal, it will still leave the result intact (it would add Marc, Althio and Janko if they haven't voted and bring the result to 11:8). However, if a better solution is proposed, I'll be happy to go on and vote for deprecating the current tag and introducing a better one. That was on the process. Now, to the actual discussion: 5. If I understand right, the main concern of the water_tap opponents is the conflict between man_made=water_tap and amenity=drinking_water. I wonder why no concern is raised about the drinking_water key. It provides the full "functionality" of amenity=drinking_water and more (since it allows the "no" and "conditional" values as well as the "legal" subtag). So there is a direct conflict but I haven't seen any proposal to deprecate drinking_water=*. 6. I find amenity=non_drinking_water a poor solution in general: it implies that the mapper knows that water is non-potable. This is not always the case. Not only it may not be known (marked); people may have different attitude to the same kind-of-potable water source. Non_drinking_water also doesn't indicate whether the water may be made potable. Note that this is asymmetric to amenity=drinking_water, which is *always* potable. 7. Personally, I believe drinking_water=* is a much better solution than amenity=drinking_water: 7.1) The source of drinking water (which, I fully agree, is important for a lot of users) may not be a dedicated amenity, and still be very useful: e.g. a public toilet in a well-developed country can provide access to drinking water, but it's not an amenity=drinking_water, it is amenity=toilet. Marking one thing with two amenity nodes is possible but (1) it's a workaround rather than a nice solution; (2) I think many people, especially tourists from less developed countries, may not even understand such tagging and will be looking for a dedicated amenity. 7.2) Drinking water may come in a huge variety of forms, for many of which there are dedicated tags. If you care about water-deprived tourists or NGOs, you should also think about water_well, water_point, spring, toilet, water and landuse tags. All of them are potentially "hiding" potable water from users, and most of them are not amenities. This means that if a tourist wants to find the nearest source of potable water, all these objects should be tagged with drinking_water=yes and the map users should search for this tag rather than for amenity=drinking_water. Therefore I would start a separate discussion on how to make sure that all sources of potable water are tagged with drinking_water=yes. 8. Most importantly: The water_tap tag was initiated to solve a specific problem without causing any additional conflicts, namely to provide the means to tag water taps *independent* from whether water is potable or not. That is to map an object, for which there is currently no means in OSM at all. After some discussion and attempts to find alternative tagging, the current proposal was found to be an optimal compromise because: 8.1) it is under man_made (there was a suggestion to make it an amenity), meaning that it can be used together with amenity=drinking_water to specify the type of the source; 8.2) it is very similar in all ways to man_made=water_well (again, I haven't seen any doubts on that one), so it should look logical to mappers; 8.3) it provides good means to tag a water source where there is a dedicated amenity but potability is unclear and thus amenity=drinking_water|non_drinking_water cannot be used; 8.4) unless used together with amenity=drinking_water, it delegates specification of potability to a dedicated drinking_water=* tag. I find it pretty well developed and fail to understand why some people find it not mature. In summary, this (and the original) discussion shows that there is clearly some space for improvement in tagging. It looks like the problem is deeper than a couple of suboptimal tags. No-one was able to propose a solution that others would unanimously accept; taking into account our joint experience and smartness, I consider it an indication of the problem in the tagging approach itself. Therefore I suggest to: (1) accept the current proposal since it does solve a specific problem following the current OSM style and doesn't cause more confusion than there already exists; (2) go on and try to find the root cause for our disagreement and a good solution. Cheers, Kotya On Fri, Jan 16, 2015 at 5:42 PM, althio althio <althio.fo...@gmail.com> wrote: > It seems that Pieren and I agree on most points. > > @François > Maybe drinkable water is a very special case... but here service/use > is much more important than object/feature. The ability to find this > water on a map or from any data consumer is useful. It can even be > essential to many people from hikers and bikers to inhabitants and > humanitarian NGO where water is in short supply. > > Also consider the possibility of a open data import of geolocalised > water points. We should import them for added value even if the > supporting physical man_made=* is unknown. > > You must tag what you know and what is useful. > man_made=water_[object] is useful. > amenity=drinking_water/non_drinking_water is useful. > > Let's tag one or the other and both when we can. For me there is no > conflict or hierarchy between these two keys. > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
