splitter r241 and mkgmap r2378
Hi all, I've committed spltter r241: - added parameter --problem-report. If keep-complete is true this parm can be used to specify the output file for the generated problem list. The planet-xxx.kml files are no longer written. With r240 I started to implement a version info, but this doesn't work yet. @Steve: please can you review the build script? In the jar file I see two files: mkgmap-version.properties contains the correct svn.version splitter-version.properties has the correct name, but svn.version is "unknown" mkgmap r2378 is in the o5m_support branch. It allows to read o5m files. If you want to try o5m format with splitter and mkgmap, the tool chain should look similar to this: (download xxx.osm.pbf) osmconvert --drop-version xxx.osm.pbf -o=xxx.o5m java [JVM_opts] -jar splitter-problem-list-r241.jar [options] --output=o5m xxx.o5m java [JVM_opts] -jar mkgmap-o5m_support-r2378.jar -c templates.args or java [JVM_opts] -jar mkgmap-o5m_support-r2378.jar *.o5m A few numbers: I have a niedersachsen.osm.pbf with ca. 138Mb osmconvert requires ~ 10 seconds to write niedersachsen.o5m with ca. 173Mb splitter-problem-list-r241 a) input pbf, output pbf, keep-complete=false, overlap=2000 : 50 seconds, ~ 121 MB output b)input pbf, output pbf, keep-complete=true, overlap=0 : 117 seconds, ~ 118 MB output c) input o5m, output o5m, keep-complete=false, overlap=2000 : 22 seconds, ~ 181 MB output d) input o5m, output o5m, keep-complete=true, overlap=0 : 49 seconds, ~ 175 MB output mkgmap runs more or less with the same speed, no matter if o5m or pbf is used. Ciao, Gerd -- View this message in context: http://gis.19327.n5.nabble.com/splitter-r241-and-mkgmap-r2378-tp5736852.html Sent from the Mkgmap Development mailing list archive at Nabble.com.
On 21/11/12 15:23, GerdP wrote:
With r240 I started to implement a version info, but this doesn't work yet. @Steve: please can you review the build script? In the jar file I see two files: mkgmap-version.properties contains the correct svn.version splitter-version.properties has the correct name, but svn.version is "unknown"
I've just checked it out and built it. The --version option appears to work perfectly so I'm not sure what you mean! ..Steve
Steve Ratcliffe wrote
On 21/11/12 15:23, GerdP wrote:
With r240 I started to implement a version info, but this doesn't work yet. @Steve: please can you review the build script? In the jar file I see two files: mkgmap-version.properties contains the correct svn.version splitter-version.properties has the correct name, but svn.version is "unknown"
I've just checked it out and built it. The --version option appears to work perfectly so I'm not sure what you mean!
Please look at the splitter-problem-list-r241.jar from http://www.mkgmap.org.uk/splitter/ I assume the mkgmap-version.properties file is generated by some extra script ? I also have a problem when I build the jar on a windows machine with Tortoise 1.7.xx and the complete splitter project is checked out. In this case, I have something like this: d:\splitter d:\splitter\.svn ... d:\splitter\branches\problem-list This sub-dir doesn't have a .svn directory (only the topmost directory has one with Tortoise svn 1.7.x) I don't know if ivy allows to check the content of a file. If yes, I suggest to run svnversion.exe if it exists and check the output. It writes "Unversioned directory" or something like 241M for a versioned directory. Gerd -- View this message in context: http://gis.19327.n5.nabble.com/splitter-r241-and-mkgmap-r2378-tp5736852p5736... Sent from the Mkgmap Development mailing list archive at Nabble.com.
On 21/11/12 15:45, GerdP wrote:
Please look at the splitter-problem-list-r241.jar from http://www.mkgmap.org.uk/splitter/ I assume the mkgmap-version.properties file is generated by some extra script ?
OK got you. Should be fixed now. ..Steve
On 21/11/12 15:45, GerdP wrote:
d:\splitter\branches\problem-list This sub-dir doesn't have a .svn directory (only the topmost directory has one with Tortoise svn 1.7.x)
I guess we could just remove the check for .svn I always check out a branch separately so it always has a .svn directory even on 1.7. ..Steve
Steve Ratcliffe wrote
On 21/11/12 15:45, GerdP wrote:
d:\splitter\branches\problem-list This sub-dir doesn't have a .svn directory (only the topmost directory has one with Tortoise svn 1.7.x)
I guess we could just remove the check for .svn
I agree, committed as r242. Steve Ratcliffe wrote
I always check out a branch separately so it always has a .svn directory even on 1.7.
Yes, probably this is the better way. Gerd -- View this message in context: http://gis.19327.n5.nabble.com/splitter-r241-and-mkgmap-r2378-tp5736852p5737... Sent from the Mkgmap Development mailing list archive at Nabble.com.
participants (2)
-
GerdP -
Steve Ratcliffe