Re: [OSM-talk] Voting on voting system for proposals
On 3/18/2015 2:43 PM, Clifford Snow wrote: Since you are involved with updating the rendering, can you tell us the process to decide what should be rendered? I realize that part of it must be stylistic, but what outside influences cause you to include a tag as part of the standard rendered OSM tile? I should preface this by stating that these are my opinions, and I know other OpenStreetMap Carto maintainers look at it differently. They are also not the opinion of my employer, MapQuest, and the MapQuest Open style has different cartographic goals. There are no policies on what is rendered, and types of features are decided on a case by case basis. Normally the process of deciding to render a feature and deciding to render a particular tag are separate. You might decide you want to render bus stops, but also find that in the region you're rendering there is a GTFS feed with better data. In OpenStreetMap Carto, these two steps are more entwined. We're aiming at mappers and want to avoid additional sources of non-OSM data. A first consideration is technical. Some of the crazy relation types out there are not designed in a way that they can be reasonably rendered with a standard toolchain. If I can't figure out how to write the SQL to be able to get a data layer suitable for rendering, it almost certainly won't be rendered. I'm only interested in rendering established tags. The primary indicator of this is usage. There are some exceptions to this like national capitals, where there are only many of them. My view is that a tag should be able to obtain reasonable usage numbers on its own merits without being rendered. I also look beyond taginfo numbers to see if they are being skewed by a small number of contributors, mechanical edits, or a bulk import. We don't want to encourage difficult to consume tagging approach. This is why we will not use disused=yes. (#111) The wiki is a source I use, but just one among many. A good read is Andy's comment about changing tags: https://github.com/gravitystorm/openstreetmap-carto/issues/230#issuecomment-29238913. It is related. And of course, all of this is done in a limited amount of available time. If I decide to work on something with the style it means I'm not working on a different part of it. It's zero sum for me, and I always have more I can work on. Rendering new types of features is about bottom of the priority list for me right now. Would you render a tag without a wiki entry, or with just a proposal? In principle, if it were an established tag? Yes. It's very unlikely an established tag would not have a wiki page. How does the fact that it may be useful to specific groups, ie, cyclists which has its own style impact your decisions? I don't particularly consider the presence of specialist styles. There are styles for most topical interests these days. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Voting on voting system for proposals
I would assume that in this phase of the OSM lifecycle most new tags would start in specialist renders. For example I expect that the current discussion about campgrounds camp_site=* leading to different types of campgrounds would be rendered in specialist renders for camping first and would be rendered to more general maps once they gain momentum. This makes it important that renderers can show raw attribute tags of namespace tags they show. In this way I can see if more information is hidden behind the symbol shown. On Thu, Mar 19, 2015 at 7:20 AM Paul Norman penor...@mac.com wrote: On 3/18/2015 2:43 PM, Clifford Snow wrote: Since you are involved with updating the rendering, can you tell us the process to decide what should be rendered? I realize that part of it must be stylistic, but what outside influences cause you to include a tag as part of the standard rendered OSM tile? I should preface this by stating that these are my opinions, and I know other OpenStreetMap Carto maintainers look at it differently. They are also not the opinion of my employer, MapQuest, and the MapQuest Open style has different cartographic goals. There are no policies on what is rendered, and types of features are decided on a case by case basis. Normally the process of deciding to render a feature and deciding to render a particular tag are separate. You might decide you want to render bus stops, but also find that in the region you're rendering there is a GTFS feed with better data. In OpenStreetMap Carto, these two steps are more entwined. We're aiming at mappers and want to avoid additional sources of non-OSM data. A first consideration is technical. Some of the crazy relation types out there are not designed in a way that they can be reasonably rendered with a standard toolchain. If I can't figure out how to write the SQL to be able to get a data layer suitable for rendering, it almost certainly won't be rendered. I'm only interested in rendering established tags. The primary indicator of this is usage. There are some exceptions to this like national capitals, where there are only many of them. My view is that a tag should be able to obtain reasonable usage numbers on its own merits without being rendered. I also look beyond taginfo numbers to see if they are being skewed by a small number of contributors, mechanical edits, or a bulk import. We don't want to encourage difficult to consume tagging approach. This is why we will not use disused=yes. (#111) The wiki is a source I use, but just one among many. A good read is Andy's comment about changing tags: https://github.com/gravitystorm/openstreetmap- carto/issues/230#issuecomment-29238913. It is related. And of course, all of this is done in a limited amount of available time. If I decide to work on something with the style it means I'm not working on a different part of it. It's zero sum for me, and I always have more I can work on. Rendering new types of features is about bottom of the priority list for me right now. Would you render a tag without a wiki entry, or with just a proposal? In principle, if it were an established tag? Yes. It's very unlikely an established tag would not have a wiki page. How does the fact that it may be useful to specific groups, ie, cyclists which has its own style impact your decisions? I don't particularly consider the presence of specialist styles. There are styles for most topical interests these days. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Voting on voting system for proposals
On 19/03/2015, Clifford Snow cliff...@snowandsnow.us wrote: Requiring an accepted proposal plus good documentation sound like a reasonable policy. I would probably add, that the tag is sufficiently used, and/or be very desirable. Note that actual use is far more important than documentation. For example, some time ago a change in tagging of power stations that had gone through the wiki voting process by the book did not get considered immediately because of usage stats and perceived usefullness. The thread is worth reading in full for people interested in osm-carto decision making : https://github.com/gravitystorm/openstreetmap-carto/issues/230 ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Voting on voting system for proposals
On 18 March 2015 at 21:43, Clifford Snow cliff...@snowandsnow.us wrote: Paul, Since you are involved with updating the rendering, can you tell us the process to decide what should be rendered? I realize that part of it must be stylistic, but what outside influences cause you to include a tag as part of the standard rendered OSM tile? Would you render a tag without a wiki entry, or with just a proposal? I'm not Paul, but I can give you my view: As far as I know, we don't have a policy on which tags to include in the rendering, and there is currently no consensus within the development team on what the best policy would be. Personally I'm trying to steer towards requiring an accepted proposal plus documentation on the wiki before rendering a new tag, but I know not all of the developers share this point of view. Currently, proposals for newly rendered tags are currently discussed on a case by case base. How does the fact that it may be useful to specific groups, ie, cyclists which has its own style impact your decisions? Objects aimed at specific user groups are less likely to be rendered, and if they are rendered they will appear at higher zoomlevels. For example, we don't render fire hydrants because they are of little interest to the general public. -- Matthijs ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Voting on voting system for proposals
On 19/03/2015 2:44 PM, Clifford Snow wrote: On Wed, Mar 18, 2015 at 5:45 PM, Matthijs Melissen i...@matthijsmelissen.nl mailto:i...@matthijsmelissen.nl wrote: As far as I know, we don't have a policy on which tags to include in the rendering, and there is currently no consensus within the development team on what the best policy would be. Personally I'm trying to steer towards requiring an accepted proposal plus documentation on the wiki before rendering a new tag, but I know not all of the developers share this point of view. Currently, proposals for newly rendered tags are currently discussed on a case by case base. Requiring an accepted proposal plus good documentation sound like a reasonable policy. I would probably add, that the tag is sufficiently used, and/or be very desirable. It is interesting that developers are discussing which tags to render as well as being discussed on the tagging mail list. It seems like we should have the benefit of both discussions. While not all tags need to be, or even should be displayed, I wonder if it might knowing if a tag is likely to be rendered would have an impact the acceptance of tags. It shouldn't, but it might sway voting. Even more so, the decision by developers to add the tag to editors. I would think that having a tag supported by JOSM and iD would more quickly lead to its acceptance. Conversely, not including the tag could result in it being one of the many tags with limited use. Voting is all well and good, but it seems like we need to encourage dialog with developers to support new tags or understand why they don't think the tag is worthwhile of their time and effort. I feel that voting should be just part of the approval process. If we, the mappers, feel like a new tag should be adopted, then we should make sure that developers share our belief. I am not saying that developers need to be part of the initial dialog. We would probaby scare them off from ever taling to us again! I don't think renders will be interested so much in tags with low usage. And at the start new tags have low use thus they don't get rendered. Catch 22. So renders may not actually be too interested in the making of new tags? To me (and I'm not a render) I'd render things that had a good wiki page with some idea of how to render it, lots of use and some 'significance' to the map .. the 'significance' will depend on the render and their desired application. For example cycle maps might include 'bicycle repair station' despite the low usage. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Voting on voting system for proposals
On Wed, Mar 18, 2015 at 5:45 PM, Matthijs Melissen i...@matthijsmelissen.nl wrote: As far as I know, we don't have a policy on which tags to include in the rendering, and there is currently no consensus within the development team on what the best policy would be. Personally I'm trying to steer towards requiring an accepted proposal plus documentation on the wiki before rendering a new tag, but I know not all of the developers share this point of view. Currently, proposals for newly rendered tags are currently discussed on a case by case base. Requiring an accepted proposal plus good documentation sound like a reasonable policy. I would probably add, that the tag is sufficiently used, and/or be very desirable. It is interesting that developers are discussing which tags to render as well as being discussed on the tagging mail list. It seems like we should have the benefit of both discussions. While not all tags need to be, or even should be displayed, I wonder if it might knowing if a tag is likely to be rendered would have an impact the acceptance of tags. It shouldn't, but it might sway voting. Even more so, the decision by developers to add the tag to editors. I would think that having a tag supported by JOSM and iD would more quickly lead to its acceptance. Conversely, not including the tag could result in it being one of the many tags with limited use. Voting is all well and good, but it seems like we need to encourage dialog with developers to support new tags or understand why they don't think the tag is worthwhile of their time and effort. I feel that voting should be just part of the approval process. If we, the mappers, feel like a new tag should be adopted, then we should make sure that developers share our belief. I am not saying that developers need to be part of the initial dialog. We would probaby scare them off from ever taling to us again! Clifford -- @osm_seattle osm_seattle.snowandsnow.us OpenStreetMap: Maps with a human touch ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Voting on voting system for proposals
On 3/18/2015 2:40 AM, Martin Koppenhoefer wrote: I'd like to point to the tagging mailing list, where there is currently a discussion going on, whether the current voting system for voting proposals should be changed. Just as a clarification, this is for voting on what it takes to indicate a tag is approved on the wiki. It is not about if a tag is approved for use, as there is no such thing. No approval is needed to create a new tag, to render a tag, or to otherwise do something with a tag that has not passed a wiki vote. ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk
Re: [OSM-talk] Voting on voting system for proposals
On Wed, Mar 18, 2015 at 1:50 PM, Paul Norman penor...@mac.com wrote: Just as a clarification, this is for voting on what it takes to indicate a tag is approved on the wiki. It is not about if a tag is approved for use, as there is no such thing. No approval is needed to create a new tag, to render a tag, or to otherwise do something with a tag that has not passed a wiki vote. Paul, Since you are involved with updating the rendering, can you tell us the process to decide what should be rendered? I realize that part of it must be stylistic, but what outside influences cause you to include a tag as part of the standard rendered OSM tile? Would you render a tag without a wiki entry, or with just a proposal? How does the fact that it may be useful to specific groups, ie, cyclists which has its own style impact your decisions? Clifford -- @osm_seattle osm_seattle.snowandsnow.us OpenStreetMap: Maps with a human touch ___ talk mailing list talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk