Comments on branch http://gpsmid.git.sourceforge.net/git/gitweb.cgi?p=gpsmid/GpsMid;a=shortlog;h=refs/heads/master_extended in git
Broken Route Line (in Munich map)
Issue 1: Broken route line in Munich map, see pic1
Issue 2: Overlapping buttons * and _, see pic1
probably on new tile loading sometimes the not loaded tile greys out the complete map screen until loaded which is a very irritating effect.
Issue 4: Osm2GpsMid reports 0 bytes for connections for route tiles:
8 MB for nodes in 467 files, 0 bytes for connections in 467 files
Is in http://gpsmid.git.sourceforge.net/git/gitweb.cgi?p=gpsmid/GpsMid;a=commitdiff;h=1ec63a9630b56099dd92de772983c57ab8b734a1 the value 65636 correct?
About Issue 1:
Thats strange. I tested a new map and got a broken route line in other places than in the screen shot. This seems to depend on the emulator.
I made the int threshold value float because I had a broken route line on my mobile. My newly chosen threshold value 1.5 was to small. The old value 2.0 works better. I'll commit this. But we should think about what is the real correct value. Especially because the precision of the map can be set in the .properties file.
About Issue 2:
I like to have a small gap between the buttons. I'll have a look why the offset doesn't work correctly.
Issue 2: already fixed
About Issuse 3:
You are right. Any not loaded tile covering the map will grey out the map. I think there is not much that can be done about. I think it's about to like it or not.
The best would be to have the possibility to activate/disable this per GUI. Also thinking about older slow devices.
About Issue 4:
As far as I rembember, this happens because I disabled some code which used memory but the data in memory where never used again. I'll check this.
About Issue 1: Think it would be better to find out what the correct value is and see if it can be handled by int as float is much slower. Actually had only very few broken route line issues with current git and I don't think they are directly related to the "< 2" condition.
About Issue 5:
This was a typo. I fixed this.
About issue 1:
If the float thing fixes some broken route lines I think it is worth the extra time.
I changed this since < 2 only will come out true in case the difference is 0 or 1. So anything from 1.5 to 2.0 will be rounded away by integer calculation.
Regarding issue 1:
The only broken route line I'm currently aware of is at lat / lon displayed in pic2.png (screenshot from latest master_extended), generated with attached MucSmall.properties from an extract from bayern.osm.pbf but I guess this related to the roundabout.
Can you please give an example .properties, tell which .pbf and coordinates or a gpx file containing start and destination favorites to reproduce issue 1 with master git?
Issue 1 with all master_extended commits of 05-Apr-2012
.properties file to reproduce issue 1 / pic2.png
>Especially because the precision of the map can be set in the .properties file.
About that, if the precision setting causes extra work for real useful features, feel free to drop the precision setting rather than add extra logic for it. I added it mainly as a debug tool for splitters (line & area), and it probably doesn't have a real valid purpose, so I think it's not necessarily that good an idea to add code to respect the precision setting.
Issue 3: Maybe the not loaded tile could be painted first (in the background) to not cover the already loaded map data?