A long-simmering question I have had is what sort of advice / guidance / prescription / discouragement should be included for working with fonts that wander outside the "official" list of ten Indic2 scripts.
This does happen; it was (I suspect) fairly common for designers building fonts for related scripts before USE became widespread and reliable enough, or for targeting systems where USE was not available.
Example: the OFL-licensed Sylheti Nagari font Surma uses the Indic2 shaping features, even though Sylheti is not part of the script list for Indic2. It makes sense from a development standpoint, and I would not be surprised to find others (especially Northern-Indic scripts that might be closer to Devanagari than to the USE model), particularly recent additions to Unicode that don't have a ton of commercial font options.
Since projects like that seem to be piggybacking on Indic2, that ought to be noted somewhere. So this issue poses the question, and input is welcome.
Surma might actually be a bad example here, but I think the larger question should still be asked, eg, "always try to use USE", "expect some fonts to do odd things and roll with it", etc....
A long-simmering question I have had is what sort of advice / guidance / prescription / discouragement should be included for working with fonts that wander outside the "official" list of ten Indic2 scripts.
This does happen; it was (I suspect) fairly common for designers building fonts for related scripts before USE became widespread and reliable enough, or for targeting systems where USE was not available.
Example: the OFL-licensed Sylheti Nagari font Surma uses the Indic2 shaping features, even though Sylheti is not part of the script list for Indic2. It makes sense from a development standpoint, and I would not be surprised to find others (especially Northern-Indic scripts that might be closer to Devanagari than to the USE model), particularly recent additions to Unicode that don't have a ton of commercial font options.
Since projects like that seem to be piggybacking on Indic2, that ought to be noted somewhere. So this issue poses the question, and input is welcome.