Opened 9 years ago
Last modified 9 years ago
#12918 new enhancement
Warn about 1,2,3,4,5 year old shops/amenities
Reported by: | openstreetmap.org-user-d1g | Owned by: | team |
---|---|---|---|
Priority: | minor | Milestone: | |
Component: | Core validator | Version: | |
Keywords: | Cc: |
Description (last modified by )
- (amenity=* or shop=*) timestamp:2010-01-01/2011-01-01
- (amenity=* or shop=*) timestamp:2011-01-01/2012-01-01
PS. tag list could be extended with more POI centric keys.
PS2. We can consider additional checks for "survey:date"=* values (or if they are absent)
Attachments (0)
Change History (7)
follow-ups: 2 3 comment:1 by , 9 years ago
comment:2 by , 9 years ago
Replying to Klumbumbus:
Only 0.06% of all
amenity
also have thesurvey:date
tag.
There also many highways without surface=asphalt. It doesn't mean that specifying asphalt would harm us.
Also
survey:date
is not an established tagging schema. (The high total number ofsurvey:date
tags comes from keypadmapper3) There are also several other similar tags:check_date
,survey_date
,survey
,lastcheck
,last_checked
,updated
,...
based on what was said http://wiki.openstreetmap.org/wiki/Key:check_date it is most commonly used
data: 212.000 (10x times more than second), last users: 1367
I consider 1367 users as established usage, even this tag wasn't voted using proposal.
It's usage grows over time:
http://taginfo.openstreetmap.org/keys/survey:date - 231 881 objects and 1 565 different users
follow-up: 5 comment:3 by , 9 years ago
Replying to Klumbumbus:
So to warn only about the outdated
survey:date
seems not usefull
Well I suggested validation based on last modification date and not about any tag. As you can see, there no "survey:date" checks in my queries.
Yes, we can adapt survey:date as separate checks. I think that 1500 users would appreciate this.
comment:4 by , 9 years ago
Description: | modified (diff) |
---|---|
Summary: | Warn about 1,2,3,4,5 year old shops/amenities without recent survey:date=* → Warn about 1,2,3,4,5 year old shops/amenities |
follow-up: 6 comment:5 by , 9 years ago
Replying to openstreetmap.org-user-d1g:
Replying to Klumbumbus:
So to warn only about the outdated
survey:date
seems not usefull
Well I suggested validation based on last modification date and not about any tag.
While your ticket heading said the opposite ;)
Using the last modification date would still produce warnings for most of the checked pois. And then the POI doesn't need to be outdated just because of the old last modification date. And the POI also can be outdated despite a recent modification date.
1500 users
This number only tells you the number of last different editors of objects with this tag. The number of different users who once added this tag to the objects is lower.
comment:6 by , 9 years ago
Replying to Klumbumbus:
While your ticket heading said the opposite ;)
True that I don't know about exact algorithm how it should be implemented.
I use date tags+last modification date+authors
Mapbox members use only modificaion date:
"Find potentially stale information (or fresh edits)"
and they apply this to everything, not to just shop/amenities
https://www.mapbox.com/blog/2012-08-15-using-filters-josm/
But that's too demanding IMO. checks for outdated relations and buildings should be separate from shops/amenities.
Using the last modification date would still produce warnings for most of the checked pois.
Yes, some of the users want this. This check(s) can be disabled by default.
And then the POI doesn't need to be outdated just because of the old last modification date. And the POI also can be outdated despite a recent modification date.
Idea is that if you update object or update special tag, then validator shouldn't show a warning.
1500 users
This number only tells you the number of last different editors of objects with this tag. The number of different users who once added this tag to the objects is lower.
Yes, AFAIK we don't have public tools to show who added what and when.
Tags are not used if the no support for them in software and editors.
Yes they were added, because not many users think they would help.
I don't think that using http://www.openstreetmap.org/note/1 "there no such thing" is better than modifying objects themself.
Editors and validation software should support "I have seen this object at ... date" tag.
It should be optional because it takes additional time, but it should be possible to do.
Some plug ins may add these tags automatically because of very specific workflows they imply: http://wiki.openstreetmap.org/wiki/JOSM/Plugins/Surveyor
OsmAnd, Vespucci can add this tag to all nodes with POI tags, etc.
comment:7 by , 9 years ago
You can only use survey:date
if it was added or modified during last edit. Otherwise value of survey:date
is questionable.
To implement this check you may need to know who added/modified value of survey:date
for the last time. Because some users won't update value of this key.
Only 0.06% of all
amenity
also have thesurvey:date
tag. A missingsurvey:date
tag doesn't make the object more up to date than an outdatedsurvey:date
. So to warn only about the outdatedsurvey:date
seems not usefull in relation to the amenities without this tag. Warn about both (outdated or missingsurvey:date
) would produce a warning for 99.9% of all amenities, which is also not helpful.Also
survey:date
is not an established tagging schema. (The high total number ofsurvey:date
tags comes from keypadmapper3) There are also several other similar tags:check_date
,survey_date
,survey
,lastcheck
,last_checked
,updated
,...