Chiming in here: Optical kerning is not only useful for sub-par fonts; I find that even excellent fonts sometimes benefit at different sizes from different treatments. A book font like Garamond Premier Pro might perform excellently at text sizes Manually but if I ever want to deploy it in headline sizes or in a different medium (knockout white over a web image) having the Optical option is excellent.
I imagine that Optical kerning is a major programming task.
In the meantime, a QXP-style "User Defined" setting might be extremely useful and get you 90% of the way to Adobe. To wit, give the user a Font Kerning override database/table that's manipulated via a simple dialogue box. The commenter who asked for different spacing for the French "c’est" could manually kern instances of c’ and/or ’e. The database/table should record the user's Font Kerning overrides with the following data:
FONT(s) (text) • do allow for multiple fonts (bold, caption, extended, etc.) within a family
KERNING PAIR • allow for inclusion of multiple act-alikes like e, é, è, ê, ë on both left and right side of pair. Perhaps allow "any letter" or "any digit" options.
SIZES (positive integer) • allow for a point size range to which this would apply, including "ALL SIZES"
TRACKING MODIFIER (positive or negative integer) • how many percentage points tight (negative) or wide (positive) to track the particular kerning pair. This should probably modify the existing kerning table (standard 10 minus user 6 equals 4), not replace it.
If you allowed these tables to be XML import/export saveable, users could share their best practices for different fonts. En dashes for number ranges, for example, are often disgusting. Em dashes are often too tightly spaced if they're going to be used for a text dash. Ellipses are all over the place in terms of how people like to implement/kern them. Importing someone's best practices table should give you the option to ADD to existing font overrides, or REPLACE them. Perhaps you could also allow multiple rules and have the user specify if new rules will get top priority or bottom priority in a stack. You get the idea.
Good luck continuing to implement this exciting software package.
