On Sat, 21 Jun 2025 at 16:25, Russ McDermid <[email protected]> wrote:
> Question is, in a case like this, its fairly easy to "stitch" the stream > to the boundary, so long as its course has not altered over time. While > the semantics could be argued, its clearly an improvement to the map. > I profoundly disagree! The waterway is apt to be pulled around by folk aligning and realigning to various aerial views (which may or may not improve things, and may or may not account for inherent inaccuracy in aerial photos - I speak as an ex-culprit). The boundary is defined otherhow. Fine if the two ways overlie each other with perfect congruence, but imho stitching anything together on the map usually causes a headache for someone later - and boundaries are particularly awkward cases. > Apart from obvious changes (such as a waterway diversion) is there > anything wrong with adopting this approach, when it suits ? > Russ. > > On Sat, 21 Jun 2025 at 16:08, Edward Bainton <[email protected]> > wrote: > >> Without having the expertise to comment on the relevance, I can say there >> are plenty of instances of brooks being canalised, while the boundary >> remains at the historical centreline of the watercourse (as at YYYY, >> presumably, since the brook meanders continue to move around). >> >> Eg, https://www.openstreetmap.org/note/3904819 >> >> On Sat, 21 Jun 2025 at 15:56, Phil Endecott via Talk-GB < >> [email protected]> wrote: >> >>> [email protected] wrote: >>> > Send Talk-GB mailing list submissions to >>> > [email protected] >>> > >>> > To subscribe or unsubscribe via the World Wide Web, visit >>> > https://lists.openstreetmap.org/listinfo/talk-gb >>> > or, via email, send a message with subject or body 'help' to >>> > [email protected] >>> > >>> Daniel Hatton wrote: >>> > It looks like, on OSM, we have parish boundaries in England imported >>> as >>> > a numerical dataset from an OS OpenData product (they're tagged >>> > source:OS_OpenData_Boundary-Line). Unfortunately, the boundaries in >>> > this dataset appear to have been polygonized with a rather coarse >>> > granularity - about 15 metres. >>> >>> FYI, here's what OS say about the resolution of that data: >>> >>> Resolution >>> >>> The resolution of the coordinate system is 0.1 m. However, it is not >>> possible to calculate meaningful accuracy limits for Boundary-Line >>> data due to both the graphic nature and scale of the primary source >>> 1:10 000 scale published mapping. Such mapping is subject to >>> limited map generalisation, where an impression of the ground >>> detail is made due to the complexity of the detail and importance >>> of certain features such as roads. This means that boundary >>> alignments are cartographically represented in areas where >>> accurate positional representation would be impossible. >>> >>> Boundary-Line is derived from the basic scale of 1:10 000. The >>> relationship of boundaries to ground detail mirrors the accuracy >>> achieved by mapping against OS VectorMap Local (1:10 000 scale) >>> and, in certain cases, large-scale sources. A consequence of this >>> is that if Boundary-Line is superimposed upon boundaries in OS >>> MasterMap topographic data, variations in the two alignments will >>> be seen. >>> >>> >>> https://docs.os.uk/os-downloads/addressing-and-location/boundary-line/boundary-line-technical-specification/shapefile-tab-and-gkpg >>> >>> >>> Phil. >>> >>> >>> >>> >>> >>> _______________________________________________ >>> Talk-GB mailing list >>> [email protected] >>> https://lists.openstreetmap.org/listinfo/talk-gb >>> >> _______________________________________________ >> Talk-GB mailing list >> [email protected] >> https://lists.openstreetmap.org/listinfo/talk-gb >> >
_______________________________________________ Talk-GB mailing list [email protected] https://lists.openstreetmap.org/listinfo/talk-gb
