During the development of an extension i got notice of xrb files the first time.
They are used to support localization for property files. The question is why we
need this extra format and don't work on the property files directly.

Normally property files contains already a language identifier in the file name like
myproperties_en-US.properties

These files contains most often only key=value pairs and maybe comments. It
should be straight forward to parse these files instead of the one xrb file (xml
file) and to create additonal property files for new localized languages.

From my point of view we don't need the extra xrb format and we would eliminate
a further error source. Because i stumpled over a wrong encoded unicode
character. Correct encoded in the xrb file but wrong in the generated property
file (ok it's already fixed). And in case of a NetBeans project i have both
files because i want to build my extension still with NB and not only via
command line. This is redundant information and even error-prone.