[Indic] Remove Gurmukhi Yakash (and CONSONANT_MEDIAL) from spec?

[adrianwong]

Our Gurmukhi spec mentions the "Yakash" consonant, which Uniscribe apparently doesn't treat specially, and HarfBuzz relatively recently removed support for.

Should we follow suit?

If we'd like to leave it in, should we explicitly specify what its position tag should be? Considering it is positioned in a "mark-like, below-base form", should we have it tagged as POS_BELOWBASE_CONSONANT?

[behdad]

Wait on this one. We are going to revert my change.

[behdad]

Yeah that was reverted. See: https://github.com/harfbuzz/harfbuzz/issues/1592

[adrianwong]

Thanks @behdad. Interesting that Uniscribe and CoreText don't form a dotted circle if a "Yakash" comes after a matra. I wonder if they categorise "Yakash" as a matra, or if their regex permits consonant medials to occur after a matra group.

Considering the L2/06-037 proposal states that the "Yakash" should be treated as a subjoined form of "Ya", it seems sensible to me that it should be tagged as POS_BELOWBASE_CONSONANT, but I'll defer to @n8willis on this.

[behdad]

@dscorbett

[lianghai]

POS_BELOWBASE_CONSONANT feels okay.

The Unicode Core Spec already states “[t]his character should occur after the consonant to which it attaches and before any vowel sign”, and I feel it’s unlikely this character will ever get a different recommended position in the encoded cluster.

[adrianwong]

Thanks @lianghai! Good to have some confirmation on this matter.

[n8willis]

Updated to specify POS_BELOWBASE_CONSONANT in c5364de.