On 5/29/2012 1:09 AM, Alan Mintz wrote:
I used to agree with you, but in terms of minimum labor, updates are
best performed by retaining the original upload data, then doing a
conflation between the original data and a later update. That will
highlight only changes from the original source, and only those
differences will need to be manually merged into OSM.
Except you won't see possible errors introduced after the first import
by OSM editors. I think it's useful to see the diff between the current
state of both databases.
In an ideal OSM world, those errors would be caught by the
'Gardeners' in the area who tend their regions by watching OWL or an
equivalent edit monitor. The best time to catch errors is while they
can serve as a learning experience for a new contributor who can
remember what he intended to do, as well as easier to revert if necessary.
Doing a diff between the updated database and the OSM database calls
out many changes that shouldn't need to be reviewed: a fence terminating
at a building, gardens, plazas, sidewalks and stairways that connect to
buildings. It's just a trade off in the effort needed to perform the
import synchronization task.
_______________________________________________
talk mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/talk
