#9682 closed defect (fixed)
[Patch] Unneeded keys created by JOSM
Reported by: | naoliv | Owned by: | team |
---|---|---|---|
Priority: | normal | Milestone: | 14.05 |
Component: | Core | Version: | |
Keywords: | Cc: |
Description
I was validating some data that had cutting=no
+ embankment=no
+ tunnel=no
and saw that they were created by JOSM.
If you go to Presets → Highways → Streets → Residential, for example, you can see that all the check boxes allow a ternary state (yes, no, empty).
Except oneway
and lit
, the other options should only allow yes
or empty (key=no
shouldn't be created for them)
JOSM should also validate and remove such keys (like it alreay does to bridge=no
).
Attachments (2)
Change History (16)
comment:1 by , 11 years ago
Resolution: | → wontfix |
---|---|
Status: | new → closed |
by , 11 years ago
Attachment: | unclear.png added |
---|
comment:2 by , 11 years ago
Yes, I agree that =no
isn't the same as no key at all, but only for some keys. oneway
is an example that can be used to force a motorway to be two-way or also used to say "I have surveyed this street and I am sure that it's not oneway"
But the problem that I see is that people don't intentionally specify that a street in the middle of a city is cutting=no
+ embankment=no
+ tunnel=no
+ bridge=no
but they just unclick the check mark. We have keys with no
just because JOSM's interface allows (and easily creates) them.
It seems to me that the problem is how this ternary state is represent in JOSM. For example:
How will new users know that the state of bridge and cutting are different, since both are unchecked? He wrongly clicks on "bridge", see that it's wrong and then unclick (since the street isn't a bridge). Now he has a bridge=no
and doesn't know that this key is being created.
If you change this check box representation to a drop-down list having "yes, no, empty" I see that it will make things worst: people will start tagging everything with "no".
Could at least cutting, embankment, tunnel and bridge have the check box changed to "yes" and "empty" only? Advanced users could still add the key manually if they want (I still can't see a reason for this, however; it's the same as adding highway=no
to every sports pitch, since they aren't highway)
comment:3 by , 11 years ago
Milestone: | → 14.02 |
---|---|
Resolution: | wontfix |
Status: | closed → reopened |
Let's keep it open for a discussion …
comment:4 by , 11 years ago
Milestone: | 14.02 → 14.03 |
---|
comment:5 by , 11 years ago
Milestone: | 14.03 |
---|
comment:6 by , 11 years ago
Better documentation in the wiki would definitly help. I did add a note but could not find the different checkbox images in source to use them in the wiki. (any hint ?)
comment:7 by , 11 years ago
It would help, yes. But couldn't the interface also be simplified to only "yes" and empty for some values? (bridge, tunnel, cutting and embankment)
comment:9 by , 11 years ago
Milestone: | → 14.05 |
---|---|
Summary: | Unneeded keys created by JOSM → [Patch] Unneeded keys created by JOSM |
by , 11 years ago
Attachment: | 9682.patch added |
---|
follow-up: 14 comment:12 by , 11 years ago
Thought there was another ticket about it. Needed enhancement !
Did you intentially remove the motorroad=yes
in line 458 ?
An new chunk for the four highway related tags might be smoother which would be also useful for some railway=*
.
comment:14 by , 11 years ago
Replying to skyper:
Did you intentially remove the
motorroad=yes
in line 458 ?
Good catch. Of course not :-)
An new chunk for the four highway related tags might be smoother which would be also useful for some
railway=*
.
Hm, currently, chunks are not supported to be used inside <checkgroup>
s since chunks may contain arbitrary items and checkgroups expect <check>
s only.
The value "key=no" is not equal to "key not existing". The second can also mean, that nobody cared for that key. So if someone want to specify these keys we wont forbid it. If you feel that this has been overused, than contact the mapper and talk with him about proper usage of ...=no.
This is no software issue, but a usage rules issue. And yes we remove some keys. I hope only in cases, where the advantage of removing them is greater than the disadvantage of disturbing mappers.