[Talk-us] Proposed mechanical edit - remove objects that are not existing according to source of GNIS import that added them

Martijn van Exel m at rtijn.org
Fri Mar 22 14:22:36 UTC 2019


> On Mar 22, 2019, at 4:08 AM, Mark Wagner <mark+osm at carnildo.com> wrote:
> 
> On Thu, 21 Mar 2019 13:23:48 -0600
> Martijn van Exel <m at rtijn.org> wrote:
> 
>>> On Mar 21, 2019, at 12:35 PM, Mark Wagner <mark+osm at carnildo.com>
>>> wrote:
>>> 
>>> On Wed, 20 Mar 2019 21:46:59 -0600
>>> Martijn van Exel <m at rtijn.org <mailto:m at rtijn.org>> wrote:
>>> 
>>>>> On Mar 20, 2019, at 9:01 AM, Mateusz Konieczny
>>>>> <matkoniecz at tutanota.com> wrote:
>>>>> 
>>>>> I plan to run an automated edit that will revert part of the GNIS
>>>>> import that added them and delete objects that never had any
>>>>> reason to appear in the OSM database in any form, at least
>>>>> according to GNIS data.
>>>>> 
>>>>> Please comment no matter what you think about this idea! I will
>>>>> not make the edit without a clear support so please comment if
>>>>> you think that it is a good idea and if you think that it should
>>>>> not be done.     
>>>> 
>>>> 
>>>> Thanks for bringing the idea up. It actually did come up fairly
>>>> recently on Slack
>>>> https://osmus.slack.com/archives/C029HV951/p1550176430103000 
>>>> 
>>>> My view is that we would be missing an opportunity to have mappers
>>>> review these locations and update the areas concerned. These nodes
>>>> exist mostly in ‘undermapped' / remote areas that could use some
>>>> human mapper attention. So I’d be in favor of trying to resolve
>>>> this using some human driven cleanup first.  
>>> 
>>> My experience is that this will mostly just make things worse.
>>> 
>>> There was a MapRoulette task a while back for cleaning up
>>> unmodified GNIS-imported schools.  There were only a few of them
>>> left around me, but the most common result was that an armchair
>>> mapper would drag the node to a nearby non-house-looking building,
>>> trace the building, and merge it with the imported node.  Not one
>>> of these was actually a school.
>>> 
>> 
>> Do you think this could have been prevented had there been better
>> instructions?
> 
> No, I don't.  Sorting out which GNIS nodes are outdated and which are
> merely misplaced isn't something that can reliably be done from aerial
> imagery.  For something like "(historical)" GNIS nodes, it's better
> just to delete all of them.
> 

Short of messaging individual mappers, do you see a way in which MapRoulette could be a ‘better citizen’?
I’m thinking perhaps a way to ‘report’ challenges. (Not sure how that would work though.)

Martijn


More information about the Talk-us mailing list