Language fallback
Some information on identical messages in close locales:
| Language | Total messages | Of which identical to sh |
|---|---|---|
| bs | 13508 | 2695 |
| hr | 10475 | 1212 |
| sh | 5467 | – |
| sl | 14062 | 343 |
| sr-el | 13729 | 1368 |
Slovene would be a bad choice. In my opinion, it'd be best to fallback to Bosnian, Serbian (Latin) and ultimately Croatian before recurring to English, and the data seems to back me up. :-)
I thought hr before sr: messages are identical in 11.5 % and 10 % of cases respectively. I guess either may make sense.
I completely agree with Srdjan m. Just to add up, Bosnian, Croatian, Montenegrin, and Serbian are synonymous names for Serbo-Croatian, a name used by linguistic academia world-wide with the exception of scholars in each of the Serbo-Croatian-speaking countries, at least a significant number of them, who cannot reach an agreement regarding the name of their standard language. Slovene has a less mutual intelligibility with Serbo-Croatian and is always translated in Serbo-Croatian-speaking media (TV and radio), so it is better to stick to the Serbo-Croatian varieties or to recur to English.