N.Ganesan proposed a visible atomic alternative named Malayalam Chillu Sign to solve the present problems in Malayalam Unicode due to the invisible and strippable joiners.
I am trying to analyse the proposal here.
1.Only one extra codepoint is to be encoded.
2.The visibility and null-string mapping of the joiners are solved.
3.With equivalence mapping to virama, backward compatibility can be achieved easily.
4.All the existing Cils with and without independent glyphs can be generated using a two ligature sequence. (An example of Cil with independent glyph is ര് , without independet glyph is യ് as in വായ്പാട്ടു്)
5.Replaces both virama ZWJ and virama ZWNJ for creating Cils. The glyph യ് can be generated by യ cs. ര് from ര cs.
1.Doesn’t solve the polyvalent nature of RA/RRA etc
2.No proper fallback if the Cil glyph is not available in the font.
3.Many words will have different spellings, that cannot be visually identified by the user.
4.Fallback of Cil-K , a rarely used Cil whose glyph might not be included in Modern font.
5.Security problems in IDN etc.
1.The polyvalency problem in Malayalam Unicode is not limited to Cil RA and Cil LA. The prebase RA, sub-base LA , conjunct mpa etc suffers from the same .
2.The fallback is the Cil sign and I dont see any scope for challenge here.
3.This is not a new problem. It is an existing one as we use invisible joiners now, and any encoding cannot solve this problem. By equivalence mapping of virama and cs, there wont be any spelling change, though.
4.The fallback will be ka cs. So the user knows it is the Cil of Ka. If we were to encode Cil-K atomically instead, and the modern font decides to avoid Cil-K glyph, we see a square block instead, no way to understand that it is Cil-K missing.
5.By equivalence mapping of virama and cs, there wont be any new security issue.
1.Introducing a new sign alien to Malayalam.
2.Introducing one more encoding for the Cil sequences