3 ms·
Sure, and I can see that at least somewhat with pc⎻⎻̊ as a lazy "ent" being flattened, but I'd be interested as to why the o is superscripted rather than simply
by MatthewWilkes 12y ago
Sure, and I can see that at least somewhat with pc⎻⎻̊ as a lazy "ent" being flattened, but I'd be interested as to why the o is superscripted rather than simply distinguishable at the end of the word, as well as knowing if it was usual to include the 'o' when writing lazily.
Secondly, the '-̥̊'[1] character actually looks like it'd be slower than '㍶' so it's a little counter intuitive why it would happen this way.
I'd like to see references to other similar developments or precedents, rather than a three stage theory, especially when stage 2 and 3 are so similar.
[1] - This looks brilliant in a monospace font, but crap in view mode, but I spent minutes getting it right so it's staying. It's the 2nd example from TFA.