[OSM-talk] OSM map renderer - update

Etienne Cherdlu openstreetmap-L at gj0.net
Wed Apr 19 19:17:36 BST 2006


Richard
The next version of Osmarender will have the ability to specify
whether you want round linecaps or butt linecaps on the ends of roads.

The specification is already there in the rules file but a bug^h^h^h
design oversight meant that it didn't look very good if you chose any
style other than stroke-linecap="round".  So soon you'll be able to
say stroke-linecap="round" if you want Google style cul-de-sacs or
stroke-linecap="butt" if you want Multimap style cul-de-sacs.

Etienne

On 4/19/06, Richard Fairhurst <richard at systemed.net> wrote:
> Tom wrote:
>
> > On 4/19/06, nick at hogweed.org <nick at hogweed.org> wrote:
> > <snip>
> >> Another problem is when two consecutive segments in a street have a tight
> >> angle (say 120 degrees or less) to each other. At detailed scales, the
> >> segments look non-contiguous. I'll look at the applet code to see how this
> >> has been dealt with there.
> >
> > You want round caps on your lines, I think.
>
> Round caps are a good solution for the issue under discussion here, but
> it does
> remind me... please, folks, don't use round caps for the ends of cul-de-sacs.
>
> Round caps at the ends of cul-de-sacs are a cartographic 'innovation'
> popularised by Google Maps. No-one else uses them, and with good reason. Not
> only do they not look like maps should (yes, I know, very subjective), they
> tend to imply junctions that aren't there at smaller scales.
>
> For an example, here's where I live:
> http://maps.google.co.uk/maps?f=q&hl=en&q=OX7+3RD&om=1
>
> If you zoom out a couple of times, you'll see that a junction is
> implied between
> Nine Acres Close and Banbury Hill. But as the larger scale maps show,
> there's no
> junction there. This is because the round line cap is 'extending' the road
> beyond its true length.
>
> Richard
>
>
> _______________________________________________
> talk mailing list
> talk at openstreetmap.org
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>


--
Etienne
(Forgot to reply to the list again)




More information about the talk mailing list