[Tilesathome] the tilesathome design and ideas (was: Proposal: New T at H Server structure)

[Dirk-Lüder Kreie]

OJ W schrieb:
> * osmarender projection isn't quite right (when you render an image,
> it treats dLat/dy as constant when it should be more compllicated than
> that), and the smaller the area you render in one go, the less error
> you get.  with 10km tiles the error isn't particularly noticable
> (someone post stats here - I'm sure we've run some on error % versus
> tile level).  note: this was the case when t at h was first designed -
> osmarender may have improved since, and the perl version could
> certainly solve this

The midpoint of a 256px tile is less than 0.05 pixels off at z12 at 
85°N/S, and about 0.0015 px off at z17. (this depends on the actual 
latitude of the tile, and the error gets less near the equator), so 
TilesPipe would have worked, only to fail at the same point where 
tilesAtHome now fails, i.e. the Tiles getting too complex to actually 
make use of this fact.

The smalles Zoom where the error is less than half a pixel at all times 
is z9. So this is where stitching should begin at the latest for 
lowzoom, as the error pretty much doubles for every zoom from here on 
towards z1.


> * the larger the complexity of the region, the more work that the
> rendering engine has to do. 


> again, ask on this list if you have any more specific questions


-- 

Dirk-Lüder "Deelkar" Kreie
Bremen - 53.0952°N 8.8652°E


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 260 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstreetmap.org/pipermail/tilesathome/attachments/20080603/61e9b55d/attachment.pgp>