[OSM-dev] Slippy map's future

[Nick Whitelegg]

On Wednesday 04 Oct 2006 17:49, Nick Hill wrote:
> You could set up 100 machines with the current set up. You will end up
> with 100 slow, crippled slippy servers.
>
> The bottleneck is no longer the database. It hasn't been for months.
>
> The bottleneck is the lack of rendering scheduling system. The
> bottleneck is lack of intelligent tile invalidation system.
>

Nonetheless I still think there's a need for a planet-based 
"slippy-map-in-a-box". For instance, say someone wanted to set up their own 
OSM-based site covering a small geographic area, such as the New Forest or 
the Lake District - they could populate their own database with data from 
planet covering that area only. Setting up one's own OSM-based sites in this 
way could also mean that the map provider could customise the look-and-feel 
through an osmarender rules file, Freemap rules file or similar.

To that end I still intend to press ahead on developing rendering code. It'll 
be based on my existing Freemap renderer rather than the current slippy 
renderer (as it's more mature and I won't run into the sort of problems I had 
with mod-ruby on the dev server) but using imagemagick rather than GD, which 
I managed to get installed last night - this will enable antialiasing and 
smooth line joins to happen.

If combined with an "intelligent tile invalidation system" as you say 
hopefully we can have a good and redistributable rendering system.

Nick