On Tue, 25 Nov 2008, Dirk Stöcker wrote:
Nah, that is not right. It has the handling of all layers hardcoded.
Not the layers itself. The whole processing is layer independend. And this
hardcoding is required as contrary to the normal layers the lowzoom
requires different handling for different layers.
We can move the logic into the config, but is this really worth the
effort?
I think so. I have made some effort to remove layer names and URLs
from the code and was a little bit disappointed to see "caption" show
up again.
If another layer someday needs this kind of special treatment someone
has to look through the code and find all references to "caption".
I moved the layer names to layers.conf. I will checkin that as soon as I
had enough time to test it. Now the lowzoom essentially consist of four
types:
a) base layers (captionless),
b) overlay layers (caption),
c) result layers (tile, caption, captionless) and
d) non-stiching layers (caption).
For now b) and d) are the same, but no longer hardcoded.
Ok. Checked in. Feel free to comment.
The code makes still certain assumptions (e.g. that base layers are never
stitched from anything else than the same layer), but I wont fix these
without real need.
Ciao
--
http://www.dstoecker.eu/ (PGP key available)
_______________________________________________
Tilesathome mailing list
[email protected]
http://lists.openstreetmap.org/listinfo/tilesathome
