When editing in Potlatch, I regularly notice apparently random points that are on the map, which don’t have any tags attached. Are these points storing data that is not supported by Potlatch, is there any other reason that they are there, or should they be deleted?
Also, when doing some editing in Horsham, Sussex, I noticed that someone had plotted a load of points along roads, but they weren’t joined up by ways. Again, they were untagged. Any idea what this was about?
Many thanks,
Daveemtb
I think those points are left by users of (previous versions) of Potlatch as a click on the map often created a node unintentionally. Not sure if this still happens (as easily) with current versions. The nodes can safely be removed if those nodes appear to be random and don’t have any tags.
I think they may be the result of some auto-conversion process from gpx to ways. I’m not sure though, but the dots often appear over large distances and the nodes are spaced in such a way that it looks automatic. Either way, as said, they can be removed as they are doing nothing.
IMX they’re usually left by JOSM users, for what it’s worth. But you can find out by noting down the ID, then going to a URL like http://www.openstreetmap.org/api/0.5/node/123456/history/ to look at the tags.
IMX? Anyway, my early acquaintance with Potlatch resulted in lots of random nodes (which I removed with JOSM myself). I simply cannot see how users can leave stray nodes using JOSM, they’re just too noticeable for that. Richard, it’s not like I want to accuse or dismay Potlatch or anything, it’s just my personal experience….
But maybe it is now very hard/impossible to leave stray nodes with Potlatch and maybe there are careless JOSM users or maybe they were left by scripts. Looking at the created_by tags and the history will give the answer to that indeed.
When I see random lines of nodes, and untagged nodes they nearly always lack the potlatch tag. There have been a few exceptions for single nodes, but the majority seem to be from JOSM. I also can’t see how people leave stray nodes, but I am baffled to many things people have added.
I think stray nodes may be from when people move around and click, and unintentionally create nodes. Also when ways are deleted (which is way too easy) the nodes are not deleted with them. As for the lines of nodes; I think they look automatically generated usually by there equal spacing, or other none human like layout, and therefore I think these sound more JOSM like.
I’ve never actually made ‘real’ edits with potlatch so I don’t know much on it but from what I have seen Potlatch has the setting to allow you to have a play around without the edits actually uploading. This is a great feature which partly tackles the issue of direct editing. It’s easy to create random nodes in potlatch but it is also in JOSM. And it should be, because if it wasn’t then editing would be made a pain by the lack of ease. I actually think that potlatch makes it visually clearer when nodes are added, so less nodes should be left unintentionally. It’s also pretty intuitive, although JOSM’s less intuitive interface is partly because of its more complex tools. If potlatch was to display a list of nodes that have been created in ‘this session’ which haven’t been connected to any ways, as well as a list of untagged ways, then that may help draw attention to the mistake. (not that untagged ways are a mistake but people may not have intended to leave one/any.)
IMX baffled me too, is it ‘in my experience’…?
It is, yes. 
When I used JOSM way way way back it was quite easy to leave nodes around - if you drew a way then deleted it, JOSM didn’t delete the nodes in the way. Potlatch has always deleted the nodes when you delete the way. But I’ve not used JOSM much in the last year so it may have changed.
Stray nodes are pretty obvious in Potlatch, they’re big green blobs.
JOSM still does that…