[OSM-talk] T at H error

[Frederik Ramm]

Hi,

> > It is generally believed that we need to change t at h to split its work
> > into smaller chunks somehow so that they become manageable again. It's
> > just that nobody has gotten around to do it yet ;-) This error seems
> > so occurr on very complex tiles (likely one of the newly imported AND
> > or TIGER areas which are very detailed).
> 
> It would be great if osmarender could do *real bounding boxes* that is
> throw away all bits that fall out of our bbox, rather than just insert a
> viewbox command for inkscape (or however it is done).

I think we can do that even before we pass the data into osmarender.
It is a very t at h specific requirement so maybe osmarender should not
be burdened with it. On the other hand that would require an extra XML
parsing step, in addition to all the ones we have already ;-)

I'm loathe to tie loads of things to 0.5 deployment, but in this
specific case, maybe we can let this wait until 0.5 is deployed.
Because once we got rid segments, it is going to be much easier to
just subdivide all nodes in the OSM file into 4 (or so) groups and
copy each way into all groups having at least one of its nodes.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00.09' E008°23.33'