[OSM-dev] proposal to kill areas

Etienne 80n80n at gmail.com
Thu Jul 20 20:11:22 BST 2006


On 7/20/06, Nigel Magnay <nigel.magnay at gmail.com> wrote:
>
> On 20/07/06, Etienne <80n80n at gmail.com> wrote:
> > I'm not sure that the constraint that the start and end should be the
> same
> > node is either necessary or always appropriate.
> >
> > Consider that you are mapping just one side of a large lake.  The need
> to
> > join up the two loose ends creates an artifact in the database that is
> not
> > true.  They will eventually join up but you may not be able to complete
> the
> > task all in one session.
> >
>
> Wouldn't you just keep those as segments until you had mapped the
> area, and then create the way (area) at that point? Or create a way
> for 'lake edge', and gradually expand it until it was an area?


It may never happen if part of the lake shore was in a restricted area.
That shouldn't have to prevent describing the rest of it.


> Additionally, when serving up content within a bbox there may well be
> areas
> > that are only partially within the bbox.  Incidentally, Osmarender/SVG
> seems
> > to cope with this situation quite well (not sure whether this is by
> accident
> > or design).
> >
>
> Actually, it's worse than that; for a given bbox area there may be
> areas whose segments don't even intersect the bbox, but should be
> displayed rendered (for example, displaying a section of map that's
> inside a large forest area).


In that case I guess you wouldn't be able to see the wood for the trees :-)

> Etienne
> >
> >
> > On 7/20/06, Nigel Magnay <nigel.magnay at gmail.com> wrote:
> > >
> >  IMO - Sounds to me like an area is a subclass of way, with the
> > constraint that the start and end node must be the same. The API
> > should enforce this, but it doesn't neccesarily require separate
> > storage.
> >
> > On 20/07/06, Immanuel Scholz < immanuel.scholz at gmx.de> wrote:
> > >
> > > > There are only 4 areas in OSM and osmarender treats certain tagged
> ways
> > > > as an area. I figure it'd make the server simpler and make client
> area
> > > > support simpler if 'area=true' on a way made it in to an area.
> > > >
> > > > Functionally, if they were working, an area is identical to a way
> > > > already.
> > > >
> > > > Thoughts?
> > >
> > > I am not very happy with an area defined as it is currently.
> > >
> > > The problem I see is, that there are too many possibilities to specify
> an
> > > ill-formed area.
> > >
> > > I have no problem with deleting the 4 areas (probably test areas
> anyway?),
> > > skip out the area feature from the server for now and maybe reopen the
> > > brain storming for a different data structure solution.
> > >
> > > Even if the current data structure is kept, I dislike the idea of
> > > specifying an "area" as a parametrized kind of way.
> > >
> > >
> > > Ciao, Imi
> > >
> > >
> > >
> > > _______________________________________________
> > > dev mailing list
> > > dev at openstreetmap.org
> > >
> > http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/dev
> > >
> >
> > _______________________________________________
> >
> > dev mailing list
> > dev at openstreetmap.org
> > http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/dev
> >
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/dev/attachments/20060720/eb923192/attachment.html>


More information about the dev mailing list