QUESTION #030
Testing
How much Cyrillic is lost in interfaces designed in Latin?
The layout is built in English and shipped in Russian. The difference shows up not in translation but in the rhythm of the line.
QVORIX editorialUpdated
Why the question came up
A Russian word is on average longer than an English one; Cyrillic has different density and fewer ascenders. A button designed for "Save" breaks on "Сохранить изменения".
What is already known
- Some popular typefaces have no Cyrillic at all; others include it as a secondary set with a different rhythm.
- Letter spacing tuned on Latin almost always turns out tight for Cyrillic.
What is not known yet
- How large the real reading-speed loss is, as opposed to the subjective feeling.
- What share of interface bugs in Russian versions comes specifically from line length.
Current hypothesis
Hypothesis
Most problems are solved not by choosing a typeface but by width headroom: the layout is checked against the longest real string, not the average one.
Next step
Compare identical screens in two languages and record where wrapping changes hierarchy. The result is a set of rules, not a typeface choice.