Old Russian in Russian UDpos dataset

[moshemm]

Hi!
I've noticed that udpos.Russian dataset contains texts of Old Russian. E.g. ['се', 'написах', 'грѣшьное', 'свое', 'хоженїе', 'за', 'трї', 'морѧ'] - this is a sentence from a text of 15 cent. As far as I see, Old Russian texts make up 20% of a Russian udpos dataset. As a native Russian speaker I think that the udpos dataset should contain only modern Russian texts as the shared task on Russian POS-tagging showed that performance on 17th cent. texts is much lower than on modern ones and the language differs a lot.
I wonder whether these texts where included on purpose or by chance? But I really think this is not fine + the performance on this dataset without old texts would be better.

Hope for your response!