This is meant as a documentation of an evaluation. And a place to discuss possible future features to improve the integration.
My evaluation and findings
I am wondering how to conduct a structure mapping campaign to collect parking:lane information for a city like Berlin.
A good tool for this is the HOT tasking manager (Github).
Grid
However, the default grid based approach that the tasking manager uses is not a good fit for collecting data from roads, since roads will not fit into the grids nicely.
Solution: The tasking manger can use a custom shapefile for the task areas. We created https://gist.github.com/gislars/4c42cb22469d07af9748aa89047968c2 as a test case, which is a shape per road where the roads are joined by name and then cut on each intersection. This approach is OK for the most part and a bit weird on roundabouts and such. But in general, a good solution IMO.
Biggest hurdle for this: The tasking manager only allows max 1 MB filesize for those files which will be a challenge for larger areas / mapping projects.
Editor
The tasking manager allows to configure a custom editor. This completes the workflow in a very nice way:
select a task
open the parking lane editor
save
mark the task as complete
There are two to three features that are needed for this to become a suitable solution
[ ] Support a different map position and zoom parameter.
The current parameter are default leaflet-hashhttps://zlant.github.io/parking-lanes/#17/52.47906/13.42876
The tasking manager uses
https://zlant.github.io/parking-lanes/?#map=19/52.47738006422913/13.443987295031548
Some quick code checking: We use leaflet-hash via npm here which does not support custom hashes (there is an old issue). We could fork leavlet-hash or just add a redirect to our format.
[ ] Support showing the selected task area on the lane editor.
The task area is provided via URL param in xml format
&gpx=https://tasks-backend.openstreetmap.us/api/v2/projects/278/tasks/queries/gpx/?tasks=196https://tasks-backend.openstreetmap.us/api/v2/projects/278/tasks/queries/gpx/?tasks=196
This is what this looks like in the integrated iD editor:
[ ] Support extending the changeset comment with the provided comment template.
The tasking manager provides a comment template as part of the URL. It holds a hash-tag that is configured in the tasking manager project settings. AFAIK this is used to find/group and analyse changesets made by this campaign.
&comment=#strassen-berlin-nk
Background
The tasking manager allows to set a custom background image layer &background=custom:https://tiles.codefor.de/berlin-2020-dop20rgb/{z}/{x}/{y}.png.
This is nice but I think something that is more suited for the custom editor to handle, since likely more than one background layer will be needed. At least in berlin, we have great areal images per year and it's helpful to look at multiple layers to see the parked cars properly (tree leaves, building shadows, …).
Statistics
The tasking manger allows to show custom statistics, which I did not look into, yet.
There is a second tasking manager-style project and very recently the author blogged about using road data for tasks as well https://www.openstreetmap.org/user/qeef/diary/397633 – I did not look into details, yet
This is meant as a documentation of an evaluation. And a place to discuss possible future features to improve the integration.
My evaluation and findings
I am wondering how to conduct a structure mapping campaign to collect
parking:laneinformation for a city like Berlin.A good tool for this is the HOT tasking manager (Github).
Grid
However, the default grid based approach that the tasking manager uses is not a good fit for collecting data from roads, since roads will not fit into the grids nicely.
Solution: The tasking manger can use a custom shapefile for the task areas. We created https://gist.github.com/gislars/4c42cb22469d07af9748aa89047968c2 as a test case, which is a shape per road where the roads are joined by name and then cut on each intersection. This approach is OK for the most part and a bit weird on roundabouts and such. But in general, a good solution IMO.
Biggest hurdle for this: The tasking manager only allows max 1 MB filesize for those files which will be a challenge for larger areas / mapping projects.
Editor
The tasking manager allows to configure a custom editor. This completes the workflow in a very nice way:
There are two to three features that are needed for this to become a suitable solution
[ ] Support a different map position and zoom parameter. The current parameter are default leaflet-hash
https://zlant.github.io/parking-lanes/#17/52.47906/13.42876The tasking manager useshttps://zlant.github.io/parking-lanes/?#map=19/52.47738006422913/13.443987295031548Some quick code checking: We use leaflet-hash via npm here which does not support custom hashes (there is an old issue). We could fork leavlet-hash or just add a redirect to our format.[ ] Support showing the selected task area on the lane editor. The task area is provided via URL param in xml format
&gpx=https://tasks-backend.openstreetmap.us/api/v2/projects/278/tasks/queries/gpx/?tasks=196https://tasks-backend.openstreetmap.us/api/v2/projects/278/tasks/queries/gpx/?tasks=196 This is what this looks like in the integrated iD editor:[ ] Support extending the changeset comment with the provided comment template. The tasking manager provides a comment template as part of the URL. It holds a hash-tag that is configured in the tasking manager project settings. AFAIK this is used to find/group and analyse changesets made by this campaign.
&comment=#strassen-berlin-nkBackground
The tasking manager allows to set a custom background image layer
&background=custom:https://tiles.codefor.de/berlin-2020-dop20rgb/{z}/{x}/{y}.png.This is nice but I think something that is more suited for the custom editor to handle, since likely more than one background layer will be needed. At least in berlin, we have great areal images per year and it's helpful to look at multiple layers to see the parked cars properly (tree leaves, building shadows, …).
Statistics
The tasking manger allows to show custom statistics, which I did not look into, yet.
Live test project
https://tasks.openstreetmap.us/projects/278/map/?editor=CUSTOM
Related
There is a second tasking manager-style project and very recently the author blogged about using road data for tasks as well https://www.openstreetmap.org/user/qeef/diary/397633 – I did not look into details, yet
I now know how to get an account on https://tasks.openstreetmap.us/. Contact me for details.