Achavi and OSMCha behaves differently when reviewing similar changesets made by different editors - is this a concern?

[zabop]

Below I outline 2 changesets very similar in their contents performed by 2 different editors. I am confused for the reasons why some changeset reviewer software behaves differently when inspecting the two.

Level0

I have modified the position of node 2070617751 in this Changeset: 159223834 | OpenStreetMap yesterday using Level0. I pasted n2070617751 on the text input field on the top, click “Add to editor”, then specified the new coordinates in the big input field.

The change shows up nicely on achavi:

Shown are: new position of the node, old position of the node, new position of the way the node is part of. Resultmaps also seem ok. Change doesn’t load in OSMCha properly, I get:

Changeset:159223834 Map failed

Please reload the application. If it still doesnt work please refer to usage guide.

iD

I submitted a similar changeset using iD: Changeset: 159225661 | OpenStreetMap. Here too, I modified the position of a single node within a way.

Achavi:

Shown are: new position of the node, old position of the node, new position of the way the node is part of, old position of the way the node is part of. Resultmaps also seem ok. OSMCha does load the change properly.

Question

I am not saying these differences in how changesets by Level0 and iD appear in reviewer software are a problem, but they make me concerned whether I’m using Level0 correctly. In addition to that, it would be nice to be able to use OSMCha to review my Level0-made changesets, especially when I’m submitting way bigger changesets than the above ones.

Should I change how I use Level0, or are these differences not something I should be concerned with?

[ikonor]

This is a bug in OSMCha:
osmcha-frontend#615 “Map failed” when attempting to view specific changeset

OSMCha queries changeset data from Overpass API by bounding box (bbox) and time range. It seems that old ways only touching the bbox with a single node are returned without their nodes, which OSMCha doesn’t handle. In the second (iD) example, it probably works because the old way is crossing the bbox.

So not related to the editor and no worries about using Level0.

[drolbr]

Thank you @zabop for reporting and thank you @ikonor for replying.
The underlying issue should now have been fixed.

It is a good example how communication is finally working: I have not been aware that there has been a problem with changeset display at all bceause there are too related projects with issue trackers to follow them all. Once I have learned that there is an unexpected behaviour of Overpass API I have been able to identify that as a bug. The bug has been fixed in Overpass v0.7.62.4 which is now deployed.

So thank you for triggering the process that actually fixed the problem.