<html>
<body>
At 09:48 PM 23/03/2007, Artem Pavlenko wrote:<br>
<blockquote type=cite class=cite cite="">Hi David,<br><br>
On 23 Mar 2007, at 11:48, David Groom wrote:<br><br>
<blockquote type=cite class=cite cite="">It seems that the UK coastline
is beginning to render better in the Mapnik <br>
layer of the slippy map (it's a solid blue fill, rather than the odd
shaped <br>
grey polygons).<br><br>
However it does appear that the source for the coastline might be a
rather <br>
out of date planet file.<br><br>
In some places, eg <br>
<a href="http://www.openstreetmap.org/index.html?lat=50.85383936185782&lon=-1.315713476905844&zoom=15">
http://www.openstreetmap.org/index.html?lat=50.85383936185782&lon=-1.315713476905844&zoom=15</a>
<br>
,there is an exact correlation between coastline which is shown in the
<br>
Mapnik and Osmarender layers.<br><br>
In other places eg , <br>
<a href="http://www.openstreetmap.org/index.html?lat=50.81094579389664&lon=-0.09869895092239864&zoom=15">
http://www.openstreetmap.org/index.html?lat=50.81094579389664&lon=-0.09869895092239864&zoom=15</a>
<br>
,the coastline outline in the Osmarender layer seems to be much more
recent <br>
than the Mapnik layer.<br>
</blockquote><br>
For most of the world I was using original shoreline data, hence it's not
as smooth as it should. Polygons are only available for England and
Wales, the rest gets blue shoreline lines. To have proper polygons
for the whole planet :<br><br>
1. We need to agree that storing countries boundaries as polygons is way
forward.<br><br>
2. We have to decide if storing these polygons in main OSM database is a
good idea or not. I personally think it'll be better to store them
separately. <br><br>
3. Start creating polygons. I have all the tools and I'm happy to help
anyone, but it is a big task which requires community participation.<br>
On the other hand, If one person capable of fixing most of the UK
coastline it should be peanuts for OSM as a whole. If you want to see
nice blue sea and land just do it, don't wait!</blockquote><br>
[Sorry, David, I am responding to Artem's comments, not your original
issue!]<br><br>
From the perspective of a contributor and editor of OSM data rather than
technical, I'd certainly endorse storing coastline data
separately. Currently, it takes forever to get the data uploaded in
the first place, greatly slows down downloads of an editing area of
interest when I have no interest in editting the coastline, and when I am
editting it is extremely fiddly and user-error causing when it crosses
other data.<br><br>
I'd see the advantages as:<br><br>
<x-tab> </x-tab>- Gives
ability to store more than one coastline set<br><br>
<x-tab> </x-tab>- To be
able to retrieve it as a separate layer would make coastline maintenance
much easier.<br><br>
<x-tab> </x-tab>- Makes
general editting easier<br><br>
<x-tab> </x-tab>- Protects
it from mistakes when doing general editting, which will become important
if we encourage "drive by" editing of our data by the general
public.<br><br>
<x-tab> </x-tab>- A good
test bed for determining whether general separation is a good thing
(political boundaries, bus routes, specialist POI sets, ...) and for
enhancing edit and rendering tools to cope.<br><br>
<br>
Mike<br><br>
<br><br>
<x-tab> </x-tab></body>
</html>