Some more consonant stacks issues

[mahalisyarifuddin]

Defect Report

Font

NotoSansKawi-Regular.ttf

Where the font came from, and when

Site: https://notofonts.github.io/kawi/fonts/NotoSansKawi/hinted/ttf/NotoSansKawi-Regular.ttf
Date: 2022-06-30

Font Version

1.000

OS name and version

Android 10

Application name and version

Chrome 114.0.5735.147

Issues, character data, and screenshots

After creating issue #8, I revisited the Kawi Unicode proposal and discovered that the deepest attested consonant stacks are either 3 consonants + a below-base vowel sign or 4 consonants. To further test this typeface, I developed a web page that serves as a "Turing test" for a Kawi typeface. Adapted from Aksharamukha, the page contains all attested Sanskrit consonant clusters. I even asked Mr. Rajan to update the Kawi section of the site to be more Unicode-compliant, but I haven't received a response yet. I highly recommend this site for type developers working on Indic fonts.

I found several interesting issues while testing:

• The second conjoined PA/SA/SHA touches the first conjoined consonant.
  
  𑼡𑽂𑼡𑽂𑼱‌ 𑼡𑽂𑼱𑽂𑼦‌ 𑼡𑽂𑼱𑽂𑼧‌

• The behavior of REPHA is inconsistent when adding above-base vowel marks or ALTERNATE AA.
  
  

• When REPHA + consonant + conjoined RA + below-base vowel marks (U/UU/VOCALIC R) are combined, the vowel marks overlap the conjoined RA.
  
  𑼂𑼔𑽂𑼬𑼸 𑼂𑼡𑽂𑼬𑼸 𑼂𑼣𑽂𑼬𑼸

• The below-base vowel marks below the second conjoined WA are positioned incorrectly, similar to issue Second conjoined RA + VOWEL SIGN U problem #8.
  
  𑼒𑽂𑼒𑽂𑼮𑼸 𑼒𑽂𑼡𑽂𑼮𑼸 𑼒𑽂𑼯𑽂𑼮𑼸

There are additional complex issues that I cannot explain (in the subheader "4 consonant stacks" and beyond). It would be great if these issues could be resolved, but if not, that's understandable given their complexity.

Thank you.