[OSM-talk] Streetcomplete and Geographically Large Changesets

Andy Mabbett andy at pigsonthewing.org.uk
Sun Jul 25 17:16:36 UTC 2021

On Sun, 25 Jul 2021 at 17:36, Mateusz Konieczny via talk
<talk at openstreetmap.org> wrote:

> Jul 25, 2021, 15:15 by andy at pigsonthewing.org.uk:

>> You make assertions; they are not facts.
> StreetComplete explicitly ask to use it for survey and survey only.

With the possible exception of aerial imagery, all of the scenarios I
listed involve surveying.

And even then, I have used aerial imagery to confirm things I have
surveyed on foot, like the capacity of large surface-carparks.

> Please use it for direct survey - and if someone uses it in other way
> then all complication are fault of the user.

Now you are inventing the term "direct survey" - where is that defined?

> * Mapillary, Wikimedia Commons, and other public streetlevel imagery
> sources are not survey sources

They can be. Suppose I fix a dashcam, and upload the images from my
journey to Mapilary or Commons, as it is not safe to edit and drive at
the same time? In what way is that not a "survey"? Even when walking,
I use my camera as a visual notepad, and have seen other mappers do

> Again, they definitely should not be used with StreetComplete.

Again: baseless assertion.

> (or de facto survey, though making paper notes for StreetComplete
> is weird and loses nearly all of its usefulness).

It is not "weird", nor lacking in usefulness, for a number of reasons
(Examples: The user has a low battery. Or is in an area where it is
not wise to have a mobile on view. Or had no mobile signal and so
could not download quests.) Please don't assume the way you choose to
map is the only good way.

> Survey can be normal one, or some unusual way of obtaining
> accurate and recent info, but please never insert data using SC
> based on possibly outdated/wrong source (guessing, aerial,
> Mapillary made weeks earlier...)

None of the scenarios I have listed involve guessing, nor out-of-date
sources. You're making straw-man arguments.

Andy Mabbett

More information about the talk mailing list