forked from GitHub/gf-core
96f36d8f0a
It is now enough to make one server call per translation, and the results can cached by the browser.