Closed fchorney closed 7 years ago
This looks very odd, thanks for the report. What mapnik version are you seeing this with and what libpq client (postgres version)? /cc @artemp
$ mapnik-config --version
3.0.12
$ psql --version
psql (PostgreSQL) 9.3.16
That's what you're looking for?
so... I'm starting to think that maybe this is just a user error somehow. I decided to install the python mapnik bindings again, (and I ran the tests this time, I think I forgot to last time), and it seems to be working correctly now. Was using it in a virtualenv as well. Might have just been somehow half installed or something?
it seems to be working correctly now
That is great to hear.
Might have just been somehow half installed or something?
Possible yes. Duplicate versions or partial installs can cause bizarre behavior. A fully clean install is a good first step.
I close this now that you think it is working, please re-open if you hit the error again.
OK, so I'm not sure if this is because I'm multi threading or what, but I seem to be having this weird issue with my connection string being weird and corrupt.
My mapnik.Map and mapnik.load_map are both being called after the process is created, and only exists within that process.
My xml for this one spot it seems to be erroring out on looks as such [certain things have place holders as to not have actual username/password/etc
base.xml (srs= is the "line 23" that the error points to)
So what I seem to be getting at the mapnik.load_map function is as follows
As you can see, I seem to be getting weird big numbers in the host/port/dbname/user fields for some reason. They seem to be different every time. Is the memory for that stuff being corrupted somehow? I'm not sure where to go from here. Thanks