Opened 15 years ago
Closed 15 years ago
#3754 closed defect (wontfix)
Unable to connect ways when they're really close together due zoom level restrictions
Reported by: | Owned by: | ||
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | Core | Version: | latest |
Keywords: | Cc: |
Description
I used to be able to keep zooming in JOSM. Now it appears that an artificial limit has been added to how much I can zoom. This is a regression.
I've been unable to fix several problems in the data where nodes have been really close to other features due to this.
Attachments (1)
Change History (7)
follow-up: 2 comment:1 by , 15 years ago
Owner: | changed from | to
---|---|
Status: | new → needinfo |
comment:2 by , 15 years ago
Replying to stoecker:
When your scale has been much higher, then tell which data and which projection you used, otherwise this is a WONTFIX.
I was using Mercator trying to fix actual data in the OSM db. Sometimes I had e.g. nodes end so close to a way that even with maximum zoom I wasn't able select them and connect them to the way. Instead I had to resort to lame hacks like inserting another node in the way, splitting it etc.
comment:3 by , 15 years ago
Resolution: | → wontfix |
---|---|
Status: | needinfo → closed |
You can use middle click to switch selection (has been improved recently) or use a select-box to select such close nodes. There are multiple ways to select elements in cases like yours, but deep zoom has to many related problems to reactivate it. So this is WONTFIX.
comment:4 by , 15 years ago
Resolution: | wontfix |
---|---|
Status: | closed → reopened |
Summary: | I'm unable to zoom infinitely anymore → Unable to connect ways when they're really close together due zoom level restrictions |
(reopening because the message you wontfix-ed with shows I haven't managed to explain what the issue is)
The problem is not that I'm not able to select some nodes. I can select every node in the dataset just fine. The problem is when I have two ways like:
o | A | |o-B-B-B- | A | o
If I select the node at the end of way B (o) and try to connect it to A I'm unable to do so if they're too close together at my zoom level since when I'm hovering over it clicking will just select the node I have selected again.
I'm thus forced to connect it further up way A so that the ways won't be connected at a 90 degree angle and then drag the node where they're connected so that they'll again be connected at a 90 degree angle. But that'll introduce an inaccuracy since I have to drag a node in way A so it won't be straight anymore.
But perhaps I'm missing something obvious and there indeed is a way to connect the motorway in attachment:3754.osm to the trunk way without dragging a node on the trunk way later, or doing some hackery like splitting up the motorway, deleting the end of it and then trying to connect again.
comment:5 by , 15 years ago
I would do it as follows: select o of way B, press J. Then select the three nodes A, o, A and press L.
comment:6 by , 15 years ago
Resolution: | → wontfix |
---|---|
Status: | reopened → closed |
As said you need some of the more advanced selection methods.
- Draw a selection box (i.e. click in free space and drag mouse). This allows you to select one or more nodes at once. You can then use the selection dialog to reduce the number of selected nodes when you selected too much.
- Use middle click. With middle click you can toggle through nearby laying nodes. Each time you press middle mouse the next possible selection (either way or node) is choosen.
For your example: You are concerned, that line is no longer exact one direction after using "J"? Then do following: Select end node. Press "J". Select this node and the two other nodes of line and select "L" to make it straight again.
But I do not care at all for such things, as in reality at a junction the streets don't go 100% correct.
You can't zoom deeper that approx 10cm or less in the scale, which means a dot approx. has a width of something below 1cm. As our whole system accuracy is about 10-50 m, this is a valid limit (i.e. is 100 times better than our accuracy).
When your scale has been much higher, then tell which data and which projection you used, otherwise this is a WONTFIX.