I put a little work in my access map and created some new styles for
quality assurance. I tried to keep it as simple as possible, so the
drawing scheme is down to six colours and two line types:
Each of the different (most important) means of transport has its own
layer in which *only* explicit acces-Keys (e.g. bicycle=*) are
The second map uses inheritance according to the following tree:
AFAIK it's the first map to draw, what routing algorithms already
should rely on. I belive that this map will increase the quality of
routing information in OSM and therefore enhance OSM a lot.
Unfortunately, the result is 17 different styles, but they should not
change much in the future. (Hopefully, they don't change at all!) :)
The Styles for the first map are:
The styles for the second map are:
I have deleted the old access.xml and bicycle.xml because they are
integrated into this new structure. I don't know yet if there will be
other *_combined-styles; for the moment bicycle_combined is unchanged.
As soon as the styles are rendered, I will post it on the
I wrote a render script that (similar to the compositing that
hike&bike map does) does things that mapnik is not capable of.
My script is written in python and can render bboxes into
arbitrarily sized images, as well is able to serve tiles
However it is probably the utmost slow renderer there is,
it takes about 3 minutes for a simple tile. (Which is ok, because
it's experimental only at the time. :))
How can I make renderd (or tirex, or tilecache) call that
script, I found no documentation on that. I want to create
metatiles of course, too, because I have a thin lines of garbage
at the tile boundaries.
Can you please give me a hint/link on how/where to start?
Thanks a lot,
I'd like to start a new planet import using the new --hstore-column
feature of osm2pgsql. It would allow us to have a "name:" column,
containing all translations and thus allow better rendering of
I'll run it directly on ptolemy into the existing osm_mapnik database
but with a different prefix. The exact command is
time ./osm2pgsql \
--cache 2000 \
--style default.style \
--prefix planet_hc \
--host sql-mapnik \
--database osm_mapnik \
--hstore-column "name:" \
--hstore-column "addr:" \
--hstore-column "wikipedia:" \
to see what keys there are for rendering on my bicycle map, I query
the db for %cycle%:
SELECT key, count(*)
) AS stat
WHERE key LIKE '%cycle%'
GROUP by key
ORDER by count DESC
As this is rather slow, I was trying to take the %cycle% to the inner
query. But how can I trigger the keys in 'tags'? And would that be
Thanks for any hint,
We've ben running ptolemy as the maps toolserver for a while now and I
think it's quite stable. Over the past months a lot of people sent
requests for the multilingual maps that were running on cassini. Ad far
as I see things, this is a task for the live rendering system that will
drive the maps later published in our wikipedias. I feel it's time to do
the next one step in this direction and get up the live rendering system
I read that ortelius has been repurposed as webserver and cassini has
been renamed to adenia. From searching jira it seems that also adenia is
now running other tools. So: do we have a live maps rendering server? Or
should we use ptolemy, not only as toolserver but also as live rendering
As you may have seen on toolserver-l, I'm new here and I'd like to
improve OSM-data by displaying access rights on different layers (as
I'm doing here: http://access.t-i.ch/ ).
As far as I understand the discussion at , first thing would be to
apply for getting added to the OSM-Prject-Group, right?
This shall be done hereby. :)
Is there any documentation what I have to do thereafter to get my
tiles rendered at eg. http://toolserver.org/tiles/ti/<vehicle_type> or
is more or less 'copy-paste' from existing projects?
If so, I will try to set up a wiki-page with all your collected tipps
on how to use mod_tile so the next newbie doesn't have to bother you
that much. ;)
(Maybe for some little-used keys like 'hgv' I will also try to set up
 http://lists.wikimedia.org/pipermail/maps-l/2010-May/000623.html ff.
Marcus Buck schrieb:
> How much work is it? Half an hour? A day 9am-5pm? A month 9am-5pm? I do
> not even have a vague idea about the problems that need to be solved.
> Perhaps somebody can shed some light on it?
Setting up th styles it not an issue, let's say around 1 hour. Th think
I'm not sure about is, if we should do it on our development / toolserver.
My original Idea was to have the toolserver there for the people to play
around, test crazy or useful styles ans such.
The multilingual maps should go on the live system (what previously was
cassini). But there have been no pretensions to set up cassini with an
render stack. Maybe we just have to open a ticket in jira to get it done
. Once the system is set up, I'd go on importing another clone of the
database  and set up the multilingual rendering again - and that's it.
 i'M unsure about this one. Should we use ptolemy for live rendering
or should cassini carry it's own, local copy?
i started a planet import with the additional hstore column. It allows
us to retrive and search in all tags used in osm. I used a combined
hstore/column scheme, as our tests on the german devserver showed that
hstore-value-conditions are much slower than the column-conditions of
the classic database layout.
If the import is over, i'll come back with some sample queries.
The import is running on ptolemy using the following command:
--cache 3000 \
--prefix planet \
--style $HOME/src/osm2pgsql/default.style \
--host sql-mapnik \
--database osm_mapnik \
$HOME/data/planet-100505.osm.bz2 2>&1 | \
I'm planning to switch the ddiff update process to this database after
the import and to drop the old, classic database.