[OSM-talk] Worldwide non-surveyed tag edits

moltonel 3x Combo moltonel at gmail.com
Thu Jun 12 10:29:21 UTC 2014


A few comments about the curators VS creators (aka gardeners vs
suveyors/tracers/etc) divide :

* I'd be surprised if any OSM contributor fell in only one of the
categories. I'd say we all do a bit of both, so it's more an activity
than a profile.
* I dont think either activity is less important than the other.
Creating makes the data available, curating makes the data usable.
* The line between creating and curating is often blurry : how much
bing alignment is required to go from curating to creating ? If a new
street apears in the data once 'highwy=primary' is fixed, isn't that
"creating" ? We're all editing the same map, so isn't every edit
"curating" ?
* Contribution quality is not closely linked with the create/curate
split. It's more a combination of how clued in the contributor is, how
well he works with the community, and how much attention he gave to a
particular changeset.

Considering all this, it is strange that we place a higher "burden of
proof" on curators than creators.
* We never asked newbies to pop in IRC before creating their first
POI, even though they're likely to make a mess of it.
* We don't require a mailing list thread before tracing a big neighbouhood.
* Discussing with the community before creating a new tag is
considered good form rather than a requirement which, if unmet, allows
somebody who doesn't like that scheme to summarily revert the
changeset.

What does merit to be treated differently are large changesets (be it
the curator's worldwide typo fix or the creator's data import) :
* The amount of changes makes mistakes more likely (because we're just humans)
*,Correcting/reverting large changes can be a real PITA.
* The bigger the changeset, the more favorable the "uploading /
discussing" ratio, so we can afford more time discussing a big change,
even if it is of higher quality than a small one.


To sum it up: please stop treating/regarding "gardeners" differently
from "surveyors". If a distinction is to be made (and gradual levels
defined), it's between large and small edits.



More information about the talk mailing list