> * If you don't have that index, something is wrong.
That will be an error in adapting. Should be fixable.
> * re 6: it is correct that the spatial index is on the node, because that's
> what you are rendering.
But given that we need to join on `planet_osm_point` via the node ID to find
node tags, do we need this spatial index? I'm using the `planet_osm_id` index
in testing `p.way && !bbox!`.
> * because you can write everything that's needed from that table into the
> `planet_osm_highways_with_special_nodes` table, in this case the highway
> type, which you already do and the z_order which you don't.
That is true, but I was trying to keep it general so that a change in tags
considered wouldn't require a re-load, The TC rendering depends on quite a few
tags from the way (also `service` and `surface`). It makes thing less flexible
if we have to fix these at import time.
--
Reply to this email directly or view it on GitHub:
https://github.com/osm2pgsql-dev/osm2pgsql/discussions/2455#discussioncomment-16672338
You are receiving this because you are subscribed to this thread.
Message ID:
<osm2pgsql-dev/osm2pgsql/repo-discussions/2455/comments/[email protected]>
_______________________________________________
Tile-serving mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/tile-serving
