Why and how the text could be Bavarian

[oshfdk]

Oh, come on, we know that fewer than 7 glyphs and families account for over 90% of all spaces.

I thought the scheme works with glyph combinations across spaces? In which case, 41 different glyph combinations cover 90% of all hard spaces (dot spaces). It's the leftmost column of my table in the previous post.

y = vowels 
maybe o = vowels

...

These families are there to solve other problems that naturally arise in the flow of text. Some are splits of consonantcluster wich are too long, others are typical endings followed by a consonantcluster starting new word, and some are vowels followed by vowels.

This all must also be encoded, and it roughly corresponds to the range they cover.

It's not hard to approximate anything with anything if we add just enough rules. The more rules one has to add to make it work, the less plausible it gets. It still could be right, if it deciphers to something sensible, but without a decoding demonstration it doesn't look very promising to me so far.