Ray Haleblian

@insanemal said: OK that's odd. It's not building the path correctly. Gimme a bit and I'll get some debugging added so I can figure out where it's going bang

It's probably possible for @brunocastello to be made the reporter for this issue.

No worries peoples!

My source does not have a line 35. But that can sometimes be because of some kind of macro extrapolation that Python seems to do. @brunocastello when you run `git...

Type 'x' in that field, it appears and disappears. What's the intent -- editable or not?

Good suggestion. Arabic was on my mind years ago when working on the layout algorithm. Perhaps you could be a tester for RTL. ref: https://fontlibrary.org/en/font/droid-arabic-naskh ref: https://harfbuzz.github.io/ ref: http://designwithfontforge.com/en-US/Adding_Glyphs_to_an_Arabic_Font.html

As a side note, another case is traditional Japanese, which is vertical and RTL.

You're correct.

The general approach would be to negate the advances provided by the font and allow initialization of each line to be onto the opposite side of the page, all controllable...

Using the book and font here, i get mostly blank pages with a few printed characters outside the range of Arabic. Need to turn the debug drawing on to see...