>>>[I'm using Google Translate]
>>
>>
>>Seem Gooogle Translate works excellent with english-spain. English-turkish version is terrible-horrible... :)
>
>Just the minus? English-serbian is terrible*horrible, or sometimes terrible^horrible. The more common links, like serbian-croatian are in the just horrible range.
>
>Here's "what are you, people, using to wash carpets", from serbian original "шта ви, народе, користите за прање тепиха?" (which I typed in latinic, but google retyped it for me in cyrillic except the first word).
>
>into bosnian: "Šta ti ljudi koriste za čišćenje tepiha?" (you is now singular, "people" is now "humans" (but still acceptable), "washing" replaced with "cleaning" but, worst of all, "vi, narode, koristite" (you, people, use) is now "ti ljudi koriste" - where ti is singular-you but also "those", and so it's "thuse humans use")
>then into croatian: "Što ljudi koriste za čišćenje tepiha?" - where only štakavian "šta" (what) is now croatian štokavian "što" (what).
>then back to serbian: "Шта људи користе за чишћење тепих?" - back to štokavian što, still cleaning instead of washing, and carpet is now nominative singular (tepih) instead of genitive (tepiha), which sounds like a drunk tourist who had a six-months course of the language.
>
>So, just three languages, which are otherwise about 99% identical - even after 23 years of forced separate development with lots of official enforcement of inherited and newly added differences, we need no interpreter whatsoever - and the google manages to
>- lose the 2nd person and turn it into 3rd (you, people becomes them folks), washing becomes cleaning, most of the carpets are lost and only one is left.

So why didn't they use source control for the washing of their carpets? That would have solved the problem, by first shelving them before washing and then when they are cleaned the lost ones can be unshelved.