[mkgmap-dev] [i18n] Feature proposal for non-latin codepages

Tue Feb 26 08:53:24 GMT 2013

Hi everyone!

I'm struggling with i18n on my Oregon 450. To make a long story short: 
altough the map itself is OK (Japanese map with codepage 932), the 
device will not display any characters at all (not even latin ones).

I have tried virtually everything and come to the conclusion (for now) 
that the solution is not in the device or its software (at least not for 
now).

Then I had a look at the OSM raw data (OSM/XML) and found out, that 
labels are not in japanese only. Most nodes that have names have a tag 
set like this (don't know if my mail client will handle the Kanji 
characters properly):

  <node id="236578577" lat="35.7206495" lon="139.7787494" 
user="JaTrainWikipedia" uid="1008753" visible="true" version="10" 
changeset="13836660" timestamp="2012-11-11T18:09:02Z">
   <tag k="name" v="鶯谷 (Uguisudani)"/>
   <tag k="name:en" v="Uguisudani"/>
   <tag k="name:ja" v="鶯谷"/>
   <tag k="name:ja_rm" v="Uguisudani"/>
   <tag k="name:ru" v="Угуисудани"/>
   <tag k="railway" v="station"/>
   <tag k="wikipedia" v="ja:鶯谷駅"/>
  </node>

There are two tags that are particulary interesting: "name:ja_rm", the 
romanized version of the Japanese characters and "name:en", the english 
version of the name (wich might be in most, but not in all cases, the 
same thing). I suppose the binary format also contains this 
information?!

If there was an option like "--tag-names=ja_rm" or something like that 
that would allow me to compile a map containing only latin characters, I 
could have 2 maps: one for Mapsource/BaseCamp with proper Japanese 
labels and one containing the romanizations only for my device.

I would love to contribute the code necessary, but I'm less than bad at 
java coding :-(

What do you think of this feature proposal?

Regards
   Steffen