Yes, I know that Win9x and NT-based has some differences, the problem with ~1, ~2... numbering typically occurs when you create more than one file that has first 6 chars of LFN the same, e.g. dlouhy soubor 1.txt -> dlouhy~1.txt dlouhy soubor 2.txt -> dlouhy~2.txt dlouhy soubor 3.txt -> dlouhy~3.txt ...

Yes, but I belive it has something to do with national character encoding. If you install with selected english language you will not have this problem. I don't know ROS deeply but can you tell me what encoding is used for LFNs and SFNs for Czech lang.? I think that Windows use UTF16 for LFN transfo...

I don't want to get rid of SFN but I want to change behavior like some localized characters like á is transformed to illegal character ? when SFN is generated from LFN. If I created such file under Windows it transofrms á to different character Á that is valid.

Hi, after some time I tried to install ROS again, ver 0.3.16 (official install CD from 2.2.2014) and found that it caused a mess in LFN filenames on FAT16 filesystem when Czech language was selected. It looks like ROS use different character encoding on FS than Windows. There was created some files ...