[OSM-dev] Very long ways have been split (was: Status of Database Server after 0.4 Upgrade: Fragile)

[Lars Aronsson]

Kristian Thy wrote:

> More importantly (to me), why are we systematically destroying 
> data to fit into some arbitrary tool?

The lake is still there in physical molecules of water.  The sole 
purpose of trying to create a digital representation of its shape 
is to use that representation in some tool.  Not only JOSM, but 
any tool you can think of will have a problem with data format 
definitions that require completeness.  Some lakes (or roads or 
mountain ranges) are very large and we might not not be able to 
map all of it.  All of our solutions and defintions must be 
designed so that they can handle representations of fractions of 
reality, such as one third of a lake.  The definition that the 
periphery of a lake must be a single "way" is, in my mind, a 
broken defintion, and only useful for small ponds.  Currently I 
don't think we have any working tool to correctly render large 
lakes defined that way (with natural=water).  Instead, we're now 
drawing large lakes as coastlines (with natural=coastline).


-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se