RE: [EXTERNAL] Re: Is Devanagari ल्लाँ ambiguous?

[Andrew Glass]

Good question. We support Devanagari with our Indic engine, a peculiarity of this engine is that it doesn’t have a per-run feature application stage and all features are applied at the cluster stage. Addressing this is an outstanding issue. Therefore, the example cluster is a permitted single cluster in our Indic engine. It would certainly be possible to have a ligature more like Whitney's example, but that wasn't included in the plan for the Nirmala font. 

Cheers,

Andrew

-----Original Message-----
From: Richard Wordingham <richard.wordingham at ntlworld.com> 
Sent: 06 May 2020 02:46
To: Andrew Glass <Andrew.Glass at microsoft.com>; unicode at unicode.org
Subject: [EXTERNAL] Re: Is Devanagari ल्लाँ ambiguous?

On Tue, 5 May 2020 22:57:14 +0000
Andrew Glass <Andrew.Glass at microsoft.com> wrote:

> As a single cluster with candrabindu applied to the first l:
> 
> 0932 094D 0901 0932 094B
> 
> ल्ँलो
> 
> This cluster is supported in Nirmala UI:

Just to check, is the formation of the conjunct done within the cluster shaping or after the dissolution of the cluster boundaries?  This font seems to have been carefully designed so that a candrabindu within the consonant stack forces half-forms, an approach which prevents the vertical stacking seen in the example from Whitney, but prevents candrabindu on a consonant being rendered the same as candrabindu on a vowel.  (I tested the behaviour with the consonant stacks t.ra and t.ta, where Sanskrit doesn't permit candrabindu on the first character.)

Richard.