Opened 6 years ago
Closed 6 years ago
#17057 closed defect (wontfix)
Validator should not complain about restriction=no_entry relation
Reported by: | GerdP | Owned by: | team |
---|---|---|---|
Priority: | normal | Milestone: | |
Component: | Core validator | Version: | |
Keywords: | template_report | Cc: |
Description
What steps will reproduce the problem?
- Run Validator on attached sample.osm
What is the expected result?
No warning or error for relation 8124921
What happens instead?
Errors (1)
More than one "from" way found (1)
Warnings (1)
Role verification problem - Number of 'from' roles too high (2) (1)
Please provide any additional information below. Attach a screenshot if possible.
Build-Date:2018-12-01 19:52:23 Revision:14478 Is-Local-Build:true Identification: JOSM/1.5 (14478 SVN en) Windows 10 64-Bit OS Build number: Windows 10 Home 1803 (17134) Memory Usage: 1726 MB / 5461 MB (1066 MB allocated, but free) Java version: 1.8.0_191-b12, Oracle Corporation, Java HotSpot(TM) 64-Bit Server VM Screen: \Display0 1920x1080 Maximum Screen Size: 1920x1080 Dataset consistency test: No problems found Plugins: + FastDraw (34510) + apache-commons (34506) + buildings_tools (34724) + download_along (34503) + editgpx (34751) + ejml (34389) + geotools (34513) + gpxfilter (34506) + jaxb (34506) + jts (34524) + measurement (34529) + o5m (34405) + opendata (34698) + pbf (34576) + poly (34546) + reltoolbox (34614) + reverter (34552) + utilsplugin2 (34506) Last errors/warnings: - W: Update plugins - org.openstreetmap.josm.plugins.PluginHandler$UpdatePluginsMessagePanel[,0,0,0x0,invalid,layout=java.awt.GridBagLayout,alignmentX=0.0,alignmentY=0.0,border=,flags=9,maximumSize=,minimumSize=,preferredSize=] - W: No configuration settings found. Using hardcoded default values for all pools.
Attachments (1)
Change History (22)
by , 6 years ago
Attachment: | sample.osm added |
---|
comment:1 by , 6 years ago
Summary: | Validator should not complain about restriction=no_enty relation → Validator should not complain about restriction=no_entry relation |
---|
comment:2 by , 6 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
comment:3 by , 6 years ago
Resolution: | fixed |
---|---|
Status: | closed → reopened |
Hmm, that violates the generic turn restriction idea. I don't think that's a good idea.
comment:4 by , 6 years ago
The wiki documents it since years:
https://wiki.openstreetmap.org/wiki/Relation:restriction
comment:5 by , 6 years ago
Somebody unknown changed the wiki and added that incompatible change and it seems nobody noticed it:
https://wiki.openstreetmap.org/w/index.php?title=Relation:restriction&diff=1023151&oldid=981713
That modification is not justified by the implementation and the idea of restrictions. JOSM does not support such restrictions as far as I known and probably also no other tools.
comment:6 by , 6 years ago
We should check if routing engines support that behaviour or not. If not the wiki should be fixed. If they do, we need to implement it in JOSM as well (i.e. the map display of routing restrictions, the restriction preset and probably some other places. E.g. How does turnrestrictions plugin handle these?
comment:7 by , 6 years ago
OK, I only know them because I've implemented support for mkgmap years ago.
At that time it was rarely used but I think the argument is that you use one object for one feature.
See also https://forum.openstreetmap.org/viewtopic.php?id=64656
comment:8 by , 6 years ago
The rationale is clear, but this restriction relation was designed in a way that you only need to know if it is "no" or "only". This change requires that different types need different behaviour.
comment:9 by , 6 years ago
Yes, it makes things less easy. Same problem as with multiple via ways. My understanding is that JOSM should not complain when it is accepted. Another point is whether it supports the creation with presets etc.
If you don't like it at all I can revert the change. I just thought it was forgotton to implement it.
comment:10 by , 6 years ago
It is not that easy. Even if the change was unjustified, we can't simply ignore 4 years documented behaviour. We need to do a 'market research'.
comment:11 by , 6 years ago
A few minutes ago it was implemented in brouter, also because of the discussion in the german forum.
According to this post OSRM uses only the last from way and Graphhopper ignores all turn restrictions.
comment:12 by , 6 years ago
Does anybody know an easy way to get the number of existing relations of this type?
comment:15 by , 6 years ago
Sorry to be unclear: Affected are only these with more than one to and from. For a simple taginfo I woudn't have asked :-)
comment:17 by , 6 years ago
You can download them with the overpass api and use search type:relation ways:3-
I find 127 no_entry rels and only 13 no_exit rels with more then 2 ways.
comment:18 by , 6 years ago
Hmm. ATM my suggestion would be:
- revert this change here
- revert the wiki edit with an appropriate notice
140 relations are simply nothing compared to 1 million restriction relations.
comment:19 by , 6 years ago
After thinking this over again I no longer like the two types no_entry and no_exit at all. The concept seems broken. I see no need for a "from" role in a no_entry restriction, vice versa there I see no need for a "to" role in a no_exit restriction.
comment:21 by , 6 years ago
Resolution: | → wontfix |
---|---|
Status: | reopened → closed |
In 14494/josm: