[OSM-talk] Not attaching polygons to roads

moltonel 3x Combo moltonel at gmail.com
Fri Feb 21 21:29:55 UTC 2014

On 21/02/2014, Dave F. <davefox at madasafish.com> wrote:
> On 20/02/2014 22:40, Frederik Ramm wrote:
>> On 20.02.2014 23:04, Dave F. wrote:
>> What is *not* ok is one person "cleaning up" after the other without
>> actually adding any other improvement.

In cases that can be likened to a "change code indentation" commit I
agree, but...

>> I.e. if the other guy has connected the fields and the roads and you
>> have been *only* pulling them apart without contributing anything else
>> to the area in question, then you should have let them be;
> This bit I disagree with. Field or cemetery boundaries etc don't go to
> the centreline of the road. "Pulling them apart" & placing them where
> they are in reality is improving OSM by making it more accurate. Even if
> not boundary is added.

That's the crux of it. Separating the area from the road *is* an
improvement in itself (at least if you've got high-res imagery to
place the polygon more precisely). If that changeset gets reverted to
re-glue the area to the line (especially without engaging in
conversation with the orther contributor), it's a step backward.

>> This whole question is essentially a matter of taste, and you are
>> allowed to map according to your taste, and discouraged from enforcing
>> your taste for others.
> Disagree again, I'm afraid. Improving OSM's accuracy supersedes taste.

I agree with the "matter of taste" argument insofar as I dont complain
to mappers who initially glue areas to lines. It's just data that can
be improved like any other, and if it "tastes easyer" to that mapper,
it's fine. You really shouldn't force anybody to be more accurate than
they care to be. But again, if a mapper reduces accuracynfor taste
reasons, it's bad. If he ignores communication aboutnit, it's

There are many mapping alternatives that are a matter of taste or up
for debate. But I do think that this particular issue is matematically
clear-cut, it's basic geometry. See
for another writeup of my views on the subject.

More information about the talk mailing list