Interesting. So it is in fact a rendered-related issue. Since you've pointed out exactly where the problem is in the code, wouldn't it be better to just submit a fix and standardize the mapping practice on capital=[lowest admin_level of related boundary relations]? AFAIK this should only affect rendering programs (not routing nor indexing).
On Thu, May 15, 2014 at 6:31 AM, Martin Koppenhoefer <dieterdre...@gmail.com> wrote: > > 2014-05-15 2:51 GMT+02:00 Fernando Trebien <fernando.treb...@gmail.com>: > >> And some of these relations (though far from the top of the list) are >> not assigned an admin_centre role, even though the node exists. > > > > btw.: The current definition for administrative relations says that > admin_centre should be used one or no time in the relation, but what if > there is more than one admin_centre, e.g. entities where the administration > is split over 2 (or maybe more) places? My suggestion would be to change > this part of the relation definition in order to allow special cases: > http://wiki.openstreetmap.org/wiki/Relation:boundary > > >> >> >> It seems that German capitals follow the pattern capital=[lowest >> admin_level of relations in which the city is admin_centre], except >> Berlin. > > > > because Berlin has the capital=yes (because of current mapnik rules > capital=yes should be preferred over capital=2, as the style sheet only > takes account of capital=yes or not yes: > https://github.com/gravitystorm/openstreetmap-carto/blob/master/project.mml > ). > > cheers, > Martin > > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging > -- Fernando Trebien +55 (51) 9962-5409 "Nullius in verba." _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
