[Tilesathome] Zappy weirdness?

[80n]

This is because the planet contains numbers like this:
5.1588915616303e-06for nodes that are very close to the meridian.

The main API used to do this but got fixed some time ago.  It seems
planet.osm wasn't fixed at the same time.

I can do something at the Zappy end to convert these to proper numbers.

80n

On 8/16/07, Sebastian Spaeth <Sebastian at sspaeth.de> wrote:
>
> Hi Etienne and lines2curves.pl authors,
>
> I fetch data from Zappy. I noticed that around the London area, I get a
> lot of these when it worked flawlessly in other areas:
>
> These are z8 coordinates:
> [#1   0% jobinit] Doing tileset 128,84 (area around 52.050329,0.703125)
> [#1   0% lowzoom] Beziercurvehinting zoom level 8... ERROR
>   The following command produced an error message:
>   nice ./lines2curves.pl /tmp/output-26976-z8.svg-temp.svg > /tmp/output-
> 26976-z8.svg
>   Debug output follows:
>   | Math::Complex::make: Cannot take real part of 'nan'.
>
> Could it be that around lon=0 we try to divide by 0, or that a 0 is
> recognized as "NotaNumber"? It seems to work then anyway, just without
> beziercurvehinting.
>
> Is that a shortcoming of lines2curves.pl?
> Who knows that code?
>
> spaetz
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tilesathome/attachments/20070816/57b098c1/attachment.html>