Hi again,
so I got this one figured out too: It was using the wrong version of gcc (41
instead of 43). I now use an older version of osm2pgsql, which works fine
for me. I will send out a document with the whole installation notes when
I'm done.
However, here comes the next one:
dom@heinemann:/opt/mapnik> ./generate_tiles.py
render_tiles( (7.7731704008999998, 49.394822919600003, 10.234015614900001,
51.654049606599997) /opt/mapnik/osm.xml /opt/tiles/ 0 5 Hessen )
Traceback (most recent call last):
File "./generate_tiles.py", line 218, in <module>
render_tiles(bbox, mapfile, tile_dir, 0, 5, "Hessen")
File "./generate_tiles.py", line 135, in render_tiles
renderer = RenderThread(tile_dir, mapfile, queue, printLock, maxZoom)
File "./generate_tiles.py", line 64, in __init__
mapnik.load_map(self.m, mapfile, True)
*RuntimeError: XML document not well formed:
Entity 'prefix' not defined (encountered in file '/opt/mapnik/osm.xml' at
line 6687)*
My first guess was the XML compiler, but I'm using "libxml2". For some
stupid reason, the XML entity at the very beginning of "osm.xml":
<!DOCTYPE Map [
*<!ENTITY % entities SYSTEM "entities.xml.inc">*
%entities;
]>
is not being resolved correctly. Within entities.xml.inc, there is another
entity "settings" which than refers to the according "settings.xml.inc",
which includes an entity for "prefix". Hence, my XML compiler is not
resolving all these entities correctly. If I create an entity called
"prefix" in osm.xml, it works fine (it's still getting other missing
entities, however, it is the same issue).
I'm not planning on replacing all these entities with actual pieces of code
(which would be a task for someone, who killed his wife and daughter as we
say over here in Germany), so if anyone has ever faced something like this,
please speak up! Any feedback is very much appreciated!
Cheers
Dom
On Wed, Aug 3, 2011 at 1:17 PM, Dom L <[email protected]> wrote:
> Sure Rob,
>
> will do when I'm done.
>
> By any chance, I'm trying to compile osm2pgsql, but I'm getting
>
> heinemann:/opt/osm2pgsql # make
> make all-recursive
> make[1]: Entering directory `/opt/osm2pgsql'
> Making all in gazetteer
> make[2]: Entering directory `/opt/osm2pgsql/gazetteer'
> make[2]: Nothing to be done for `all'.
> make[2]: Leaving directory `/opt/osm2pgsql/gazetteer'
> make[2]: Entering directory `/opt/osm2pgsql'
> /bin/sh ./libtool --tag=CXX --mode=link g++ -g -O2 -pthread -L/usr/lib
> -lz -L/usr/lib64 -lpq -lxml2 -lz -lm -L/usr/lib -lbz2 -L/usr/lib64
> -L/usr/lib64 -lgeos -L/usr/lib -lproj -o osm2pgsql build_geometry.o
> input.o output-pgsql.o rb.o middle-pgsql.o osm2pgsql.o sprompt.o
> UTF8sanitizer.o expire-tiles.o keyvals.o output-null.o parse-primitive.o
> parse-xml2.o pgsql.o reprojection.o middle-ram.o output-gazetteer.o
> text-tree.o
> libtool: link: cannot find the library
> `/usr/lib64/gcc/x86_64-suse-linux/4.1.2/libstdc++.la' or unhandled argument
> `/usr/lib64/gcc/x86_64-suse-linux/4.1.2/libstdc++.la'
> make[2]: *** [osm2pgsql] Error 1
> make[2]: Leaving directory `/opt/osm2pgsql'
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory `/opt/osm2pgsql'
> make: *** [all] Error 2
>
>
> Any ideas what this is about? I wonder where it got the "4.1.2"
> subdirectory from since I got gcc 4.3 installed on my system. Tho if I look
> into "/usr/lib64/gcc/x86_64-suse-linux/4.3", I don't see any *.la files,
> hence, creating a symlink is not gonna help me...
>
> Cheers
> Dom
>
>
>
> On Wed, Aug 3, 2011 at 12:07 PM, Robert Coup
> <[email protected]>wrote:
>
>> Hey Dom,
>>
>> On Wed, Aug 3, 2011 at 7:02 PM, Dom L <[email protected]>wrote:
>>
>>> So just in case anyone ever faces the "ld: cannot find -lboost_python"
>>> too, you need to compile boost with "--with-python" option and make sure
>>> your Library path is set correctly!
>>>
>>
>> Glad you got it sorted - if you have a few minutes could you write up some
>> notes for SLED to add to http://trac.mapnik.org/wiki/LinuxInstallation ?
>>
>> Thanks,
>>
>> Rob :)
>>
>
>
>
> --
> Sent thru my vacuum cleaner.
>
>
--
Sent thru my vacuum cleaner.
_______________________________________________
Mapnik-users mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/mapnik-users
