On Tue, Mar 10, 2015 at 3:11 PM, sly (sylvain letuffe) <lis...@letuffe.org> wrote:
> Le mardi 10 mars 2015 22:14:45, Bryce Nesbitt a écrit : > It isn't really clear to me what you mean by that, but I don't like mixing > different changeset types of mechanical edits. It makes it harder to > understand what was the intended mechanical edition's goal. > Yes, the mixing does muddy the waters. On the other side: The mechanical edit is touching objects that had pre-existing pending cleanup. These are not edits that require local knowledge. Uploading these in one changeset is less version churn. In some cases it's typos ( http://blog.jochentopf.com/2015-03-05-new-taginfo-features-and-a-challenge.html ) in some cases it's modernizing tagging such as moving away from the "type" key used by relations (for example moving type=deciduous to leaf_cycle=deciduous). Since this is really a "semi mechanical edit", there's careful human review. -- To be clear: Initial post to the mailing list involved: *removing useless fixmes* Based on list feedback, the discussion shifted and was focused on one task: * remove denotation=cluster along with the fixme* (the fixme was in fact recognized as flag for the bad data). On the table now: * performing remaining needed cleanup on objects that will be edited anyway* The problems with denotation=cluster have been widely discussed (years ago and recently). It means "non-special tree". The original bot author does not agree with the mechanical edit, and wishes natural=tree to be reserved for special trees.
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging