Re: [XeTeX] Why XeTeX runs are non-deterministic?

2016-02-20 Thread Wilfred van Rooijen
If I remember well, there is (was ?) some kind of objective for backward compatibility, namely that all (updates of) versions of all TeX engines should produce the exact same output for a given input file. The background being that any TeX file that used to compile correctly in the past should a

Re: [XeTeX] New feature planned for xetex

2016-02-20 Thread Denis Bitouzé
Le 19/02/16 à 17h37, Philip Taylor a écrit : > Denis Bitouzé wrote: > >> Ha, ha, ha, so funny ! (I didn't understand anything ;) > > According to Mr Google, "سوزاک" translates to "gonorrhoea" > (blennorragie) ... Okay, now I understand and have real fun! :) -- Denis -

[XeTeX] Issue with Old Italic script & fontspec

2016-02-20 Thread David J. Perry
I am working with a group of scholars to implement the Old Italic block of Unicode. For this purpose I have created a font to support the Faliscan language. But I am getting unexpected behavior with the stylistic sets. The minimal example pasted at the end of this email gives the result sho

Re: [XeTeX] Issue with Old Italic script & fontspec

2016-02-20 Thread David J. Perry
Simon kindly tested my font in the SILE editor. The script and language settings and the stylistic set all work as expected there. So it seems the font is not the source of the problems I'm experiencing. Any additional insights are appreciated! David ---