My MapRoulette co-contributors and I are working on the next major version of MapRoulette. We would really appreciate your input and ideas. Below is my first question. I will add to this thread with more questions over time.
I don’t have a “release date” for the new version yet (almost certainly not this year), and in the meantime the current version will still be actively maintained.
Lastly, I would like this thread to be constructive. I am aware that MapRoulette is not without criticism and controversy. Fortunately, we already have a separate thread for that, so please head there if that’s what you came here for.
Discovery
One of the things I really want to accomplish is an easier way for mappers to find a task they are interested in quickly. This is especially important if you are new to MapRoulette and / or new to OSM.
Right now we have filters and a map you can explore, but I think there is room for improvement.
I am not saying MR must have this too, but I do believe these are good examples of how to create an immediate path for users to pick something that interests them.
If you’ve used MapRoulette, I am sure you have ideas about how to make it easier for mappers to find a task that suits their interests and experience level! I would love to hear them.
One thing that would be useful is a unified map view. Right now, if I zoom in on Spokane in the “browse challenges” map, I get a list of 23 challenges and a bunch of bubbles on the map with numbers in them, but there’s no indication of how many local items any of these tasks has. If l click on a challenge to find out more, I get kicked out to the global view and have to zoom in again to find local details; if I then click “Go Back”, there’s a significant pause while the overview map re-loads and re-zooms-in.
What would be nice is a quick way to see that, for example, “Fix Road Name Gaps” has only one task in Spokane County.
I’ll second that, I’m mostly interested in editing in my country (Sweden) and preferably in parts of Sweden I’m familiar with. So a map of all tasks, with filtering based on challenge etc., would be much appreciated.
As an example, I’ve built a tool that shows deviations between OSM and various Swedish governmental sources, which focuses on a map view with all deviations and filters for categories etc.: BästaJävlaKartan
Maybe an idea could be to let users tag a challenge as “global” or “country specific”, and people could look for specific countries only.
Also could be possibile to show what users participated in a specific challenge? This would allow the owner (and others?) to easier validate the edits, understand if the challenge is popular or active etc.
It will be good if you can add the possibility of edit and start the challenge directly in Vespucci. Vespucci is the best and most complete OSM editor for android, it will improve a lot the usage of Map Roulette via smartphone. Thanks
Hi, can a challenge be ignored so that it doesn’t appear in the “Find challenges” list and also not on the map?
If yes, I’d like to know how. And if no I’d like to welcome such feature very much. :)
In the new version, having a clearer process for handling flagged challenges would be very beneficial. Back in 2023 (in this comment on GitHub), Martijn said that a decision hadn’t been made yet on how to handle flagging – as part of the design for the new version, that should be decided.
It would be great to see improvements to the process for creating new challenges, particularly around increasing the chance that experienced, local mappers review prospective challenges and make comments or suggestions to improve them, before they are opened up to the general population of mappers to complete. There’s some discussion on this topic in these two GitHubissues.
Hi, can a challenge be ignored so that it doesn’t appear in the “Find challenges” list and also not on the map?
Do you mean as a user you don’t want to see a specific challenge show up in the search results? Not currently but a good idea to take up for the next version, thanks!
Yes, I’ve been thinking about that quite a bit, also in the broader sense of how to rate the quality of challenges by MapRoulette users. One idea I am considering is to enable “star ratings” ( to ) that would tell other mappers how the community rates the quality of a particular challenge. There are some issues with an approach like that, most notably:
Star ratings tend to skew towards the extremes.
Without requiring a comment when a user rates, it’s hard for the mapper who created the challenge to learn and improve.
Sorting needs to be smarter than just by star rating. Is a challenge with 5 five star ratings and 25 four star ratings “better” than a challenge with just 5 five star ratings?
If well implemented, it could help mappers and authors to make the best out of the platform. So definitely something to consider and I welcome discussion about this topic.
As for the flagging process, ideally I would like this to be a peer-reviewed system, but in the past, when I’ve asked, nobody has been interested in stepping up to participate. One solution could be to provide a challenge review environment to more experienced users (but by what measure?) where they can interact with the challenge providers.
There’s a lot more to be said about ensuring a high quality of tasks, and I don’t want to try and capture that all in one response here, but if you have constructive ideas about this topic, I very much welcome them. I also encourage anyone interested in this topic to peruse the MapRoulette Critique thread. While not always constructive, it does offer some good ideas and criticism of the current approach.
not sure whether they would be interested, but in past when I reported ongoing distributed damage caused by MR to DWG they responded that they have no ability to stop MR tasks.
Maybe create list of users allowed to suspend/archive MR tasks and list DWG members there?
But not sure whether DWG would be interested in effectively moderating MR or would they expect MR project to handle bad tasks on its own.
Different user permission levels would certainly something I would want to look into. Right now we only have regular users and a “super user” who can manage everything (I use this account to disable / delete challenges based on reports)
I could see a future where an intermediate level exists that would represent a “challenge reviewer” with elevated permissions to inspect and manage any, or perhaps just flagged, challenges.
It’s not as straightforward as just adding a column to the database though. There would need to be a process around it that requires careful consideration (who gets these permissions, what recourse do authors have, etc etc)
To help me understand what we should prioritize for MapRoulette v4, I have created a community survey. If you have used MapRoulette recently, I would really appreciate your participation!
Whats the purpose of a “5 star rating of the challenge”? I mean, whats the benefit? Even a bad rated Challenge is still online and will get keep beeing online.
I still think Voting for new tasks/challenges before they get online · Issue #2564 · maproulette/maproulette3 · GitHub would be better with a ‘simple’ upvote and downvote system.
Even with a mandatory comment for downvotes. And with the possibility to create a version 2 of the same challenge to improve.
And instead of programming your own commenting-system for each challenge https://github.com/maproulette/maproulette3/issues/2563 this should be implemented. I think it would make sense to ask the forum team if they are willing to provide an extra “maproulette”-section in the forum where you could create these threads. Dedicated to maproulette. One place for each mapper to find new challenges. To discuss them with OSM account.
You are probably right that a 5 star rating may not be the correct approach.
We have to strike a balance between making it easy to put tasks out there for mappers who want to collaborate on solving specific issues in OSM data on the one hand, and creating a system of feedback that is effective at weeding out bad actors and powered by community. A voting system would be something to consider, but my (admittedly limited recent) experience with the existing voting on tags on the OSM wiki gives me pause, because it’s fairly easy to manipulate when only a small number of mappers participate in the discussion. But maybe the other part of your comment could help solve this.
Like I commented on that ticket, I do not dislike that idea. If the forum can be set up so that it can have a section for MapRoulette challenge specific discussion that is easy for those who are not interested to ignore, I’d probably prefer that over “rolling our own”. I’ve never been really excited about having an internal commenting system when there are already so many channels available to the community, but when we implemented it, the forum did not exist yet.
I once participated in a challenge to update old (no update in OSM data for many years) restaurants. I was surprised to find a task about a restaurant updated only a month ago.
It turned out, the challenge was created before this (more or less) recent OSM update and the tasks did not update automatically. The challenge creator updated the data manually after my notice.
I guess such “update old osm data” challenges exist a lot. It would be great if those challenges update automatically to not cause confusion.
I like that a lot - I think if you did that and maybe even considered one more level for total newbies, it’d be great. In that new user tier, they could create private challenges and share links with people here or their friends, etc, but can’t create discoverable challenges until after they’ve had at least one of their challenges reviewed by someone in that new tier you’re discussing, that feels like it could create a user education pipeline for new challenge creators, and also quick recourse.
Even if the intermediate tier people couldn’t totally close a challenge, but could force it to not be discoverable immediately while any reports and discussion are resolved, I think that’d be big - both of those, in my opinion, would make MapRoulette a really great environment for welcoming new users, increase the quality of the challenges, and improve the areas where I’ve read concerns.
I know that still doesn’t address the “who does the work” side of things - since OSMUS supports development, could it also consider supporting the recruitment of a rotating set of reviewers in a semi-official capacity?
I think I remember that case @Nielkrokodil ! Auto-updating challenges is definitely something I would like to address in the next version. Right now, refreshing is possible but there is no single way to do it because tasks be sourced in three different ways (not taking into account challenges created and maintained entirely using the API). For Overpass based challenges, authors can trigger an update manually in the admin interface. For GeoJSON based challenges, you can upload a new file (if you created the challenge with an uploaded GeoJSON) or replace the file at the source and trigger a refresh if you created the challenge with a remote GeoJSON URL.
Another thing we’ve done is automatically making challenges hidden if the tasks have been created more than 6 months ago. As a challenge author, you need to go back in and refresh the challenge (using the methods above) before you can make it visible again.
For MR4 I do want to take this a step further. One idea I developed a while ago is a “continuous” challenge type (may even become the default?) that re-executes the source overpass query every so often to add new tasks and remove ones that have been fixed outside of MapRoulette. If you have further thoughts about this, please share.
to