Opened 10 years ago

Closed 10 years ago

#1300 closed defect (fixed)

upgrade of postgis/mapnik broke gpsdrive

Reported by: hamish Owned by: live-demo@…
Priority: major Milestone: OSGeoLive7.9
Component: OSGeoLive Keywords: mapnik, postgis, gpsdrive
Cc:

Description

Hi,

gpsdrive doesn't start because legacy_minimal.sql called from load_postgis.sh failed to load because it doesn't exist anymore on the disc.

see also

https://bugs.debian.org/699079 https://groups.google.com/forum/#!topic/mapnik/e9xO1JiqFv4

is it valid to copy in the old version of legacy.sql from the 7.0 disc?

Also I notice that the python-mapnik package (2.2) isn't installed, but the python-mapnik2 (2.0) version is. Is that intentional? Libraries for both versions are installed (which is fine).

postgis is version 2.1.1. postgresq is mostly version 9.3.2.

thanks, Hamish

Change history (6)

comment:1 by hamish, 10 years ago

Hi,

in the 7.0 live disc legacy.sql and legacy_minimal.sql were provided by the postgresql-9.1-postgis-2.0-scripts package and located in /usr/share/postgresql/9.3/contrib/postgis-2.1/

note that it's not just the osm_local DB, load_gisdata.sh is also trying to load it for the natural_earth2 DB.

I committed the ST_ api updates for gpsdrive back in January, maybe it just needs a rebuild?

anyway, copying in legacy_minimal.sql from the 7.0 live disc and changing paths from 2.0 to 2.1 in the .sql file works if all else fails.

Hamish

comment:2 by hamish, 10 years ago

Hopefully fixed with r11103. Kosmo and gvSIG likely also were affected by this.

Hamish

comment:3 by kalxas, 10 years ago

Can someone please confirm the fix in beta3?

comment:4 by kalxas, 10 years ago

Unfortunately, gpsdrive crashes after 2-3 secs.

comment:5 by hamish, 10 years ago

apparently the copy of legacy*.sql from gisvm/app-conf/ in load_gisdata.sh failed; should be an easy fix.

Hamish

comment:6 by kalxas, 10 years ago

Resolution: fixed
Status: newclosed

Fixed in build [11163]

Note: See TracTickets for help on using tickets.