I am excited about this! I downloaded the preview for Android and fired it up. I’ll try and share more feedback but here’s some initial points:

• “Save” suggests that you’re saving locally. Instead this commits to OSM.
• It connects to the OSM dev server which makes it hard to really try it out.
• It wasn’t immediately obvious to me if any presets are available for POI types. It just asks for a common name and it looks like you can manually add tags.
• The zooming / panning has some inertia that doesn’t feel right to me.

Observations made without reading the manual :)

@Glassman someone was interested in reviving the groups branch..It’s very stale at this point so I don’t know how feasible it is. But it’s vaguely related to what you (and I) would love to be able to do.

The original idea for the branch was to have a lightweight implementation of opt-in groups. Anyone could form a group and ask people in or give them the URL to join. The group itself could consist of a variant of the user diaries page.

The big hurdle for building groups out beyond that is spam. It’s already such a big problem with the diaries. Ideally you’d be able to ‘invite’ people within a geographical scope (or local mappers through an API like this) to join, but how do you prevent folks from using it as a spam tool? A reputation API would be handy for that :D

@zverik sourcehut is something I’m trying out. It doesn’t have search so you have to know about a repo to discover it. It’s a paid service (optional right now) but it’s independent and open source. See https://lists.sr.ht/~sircmpwn/sr.ht-announce for project updates. I don’t know if I will put more projects on there.

Flask-restful is supported by some major company so it’s pretty well maintained unlike some other Flask extensions. And as you see it lets you stand up a prototype REST API very quickly. It wouldn’t be a huge leap to get it to production level.

@Rory I’d be interested in seeing some of your workflows. That sounds really interesting.

@Bryce I never noticed that feature! Awesome. It looks like this is only fellow GoMap edits?

@Istvan @pangoSE @GoWIN thanks!

I just wanted to demonstrate something useful that is a) easy to implement and b) there for anyone to improve upon. Mateusz pointed to one difference with the Neis tool, but perhaps someone who knows their secret sauce can weigh in.

Carnildo – I think MapRoulette users can be educated / directed to be more careful. A lot of it is in the quality of the instruction, adding the proper links / warnings / caveats. I am sorry if it has caused trouble for your local area before. (Speaking as the maintainer of MapRoulette.) In the future (or even now if the challenge still exists) feel free to contact the challenge owner / author to let them know that you’re unhappy with the work mappers are doing based on their challenge.

I did actually work on a hopefully thoughtfully formulated MapRoulette challenge related to the topic of this entry, and I would appreciate some feedback on the instruction and description, especially from a MapRoulette sceptic like yourself :)

hrek:

1. You can access them through the API, like this curl -X GET "https://maproulette.org/api/v2/tasks/review?startDate=null&endDate=null&limit=10&page=0&order=ASC" -H "accept: application/json", see also the API docs

2. We’re working on refining that based on user feedback, there’s a Github ticket for it.

Thanks for the tip re the geography type cast, that is clever.

POIs mapped as ways would be good to include, but would make things a little more complicated. Fortunately you don’t need the actual way geometries which reduces osmosis load time in PG. You would need to filter by closed ways only, I don’t know if osmium lets you do this?

Hm interesting. I couldn’t find anything about it in the Utah traffic code. In any case I’m not that worried about it, but thanks. Note that in some U.S. states it is also illegal to attach anything to the windshield when it could obstruct the view.

It looks like there are no unfixed missing roads tiles in the Philippines currently! (Good work :))

GOWin, sorry I was not able to do this for you yet, are you still interested?

Hendrikklaas – your OSM profile page shows location for other mappers around you as well, you are right. But these are not based on actual mapping activity, but rather on where users define their location on their OSM profile. This is not necessarily where they map, and it’s optional so many mappers don’t set it at all.

pitscheplatsch, the tool you linked to is fantastic as well, but has a different outcome and objective. By default, it uses an ‘activity center’ (I am not quite sure how it is defined, I don’t think the tool is open source) to identify mappers active in the area. If someone is an active mapper but is active in many different places, they may not show up in this map. Also, the tool does not make it easy to get a list of local mappers. MYM was really designed to get a list of local mappers that you can sort based on the criteria that are important to you, and give enough detail to give you a more complete idea of who’s active in your area. I hope the tools can work together to give mappers the information they need!

Came here through Andy’s recent blog post https://blog.gravitystorm.co.uk/2018/11/09/moderation-authorisation-and-background-task-improvements-for-openstreetmap/ . Thanks for the update with these numbers. Curious but I guess not unexpected that there is abuse of this system as well. We’ll never completely rid ourselves of bad actors. At least moderation makes it possible to deal with the more visible results of their actions. Thanks for all the work on this!

I am glad you are finding it useful!

A couple of minor corrections are in order:

1. Because MapRoulette is not yet served over SSL, it cannot ask for browser geolocation yet. Right now it relies on the home location from your OSM user profile.
2. The ‘Intersecting Map Bounds’ check does not actually check for tasks within the map bounds (that would be expensive). It checks whether the Challenge’s computed bounding box intersects with the current map bounds. So it is not guaranteed that there are any tasks for the Challenge that fall inside the map bounds.

If we get into definitions. I can play that game too. Listening you can do passively (which is where we are at in OSM, at best) and actively. Actively listening would include, among other things, giving people a place where they can talk about difficult topics privately and confidentially. This, and having an anti-harassment policy are common (and in some jurisdictions required) in larger organizations.

If we keep saying there isn’t an issue, we keep making it harder for people who experience harassment / discrimination to speak up. Just look at how long it took for workplace harassment to be taken seriously. And there’s probably still people out there who deny that this is a problem as well. Like Mikel says - the least we can do is listen.

Carnildo, yes, I have seen many of those as well. Not everywhere though.

Also :/ it only seems to count trunk_link… Someone with more Overpass QL skills please say something useful :D

Additionally, this query counts on- as well as off-ramps. Looking at network connectivity would fix that. Perhaps this query is more easily / satisfactorily solved in PostGIS or some other geospatial database.

Rob, from http://overpass-turbo.eu/s/zmo it looks like you have 11803 motorway and trunk link ways, of which only 282 have destination. This is not a completely accurate picture of the situation (since typically only the first way connecting to the motorway / trunk has destination) but it should give you an idea :)

I don’t know if there is an overpass query that only considers the _link ways directly connected to their motorway / trunk counterparts.. Perhaps mmd can say something about this!