Burmese Rendering (dots and circles)

[Richard Wordingham]

On Fri, 2 Sep 2022 04:38:53 +0000
James Kass <jameskass at code2001.com> wrote:

> (Hi Richard, this was sent to the Unicode list, but was rejected as 
> spam.  Feel free to reply on the public list, privately, or not at 
> all.   -JK)

> On 2022-09-01 8:08 PM, Richard Wordingham via Unicode wrote:
> >> As Richard has pointed out, the <redacted> font does not support
> >> the Virama+Wa sequence.  
> > This is consistent with technical note UTN-11 Version 4, which gives
> > the encoding for Sanskrit <ktv> as <U+1000 MYANMAR LETTER KA, U+1039
> > MYANMAR SIGN VIRAMA, U+1010 MYANMAR LETTER TA, U+103D MEDIAL WA>.  
> 
> https://www.unicode.org/notes/tn11/UTN11_4.pdf
> 
> That sequence is shown on page 28 of the document.
> 
> I apologize for my inability to understand this.  Since U+103D is 
> supposedly for modern Burmese, and the "virama plus" sequences are 
> reserved for historic purposes, I expected the sequence for Burmese 
> (Sanskrit) <ktv> to be 1000 1039 1010 1039 101D [ka + virama + ta + 
> virama + wa].
> 
> It's also not clear to me how any given font's lack of support for 
> rendering a "virama plus" sequence is consistent with UTN11.

> Should a Myanmar Unicode font support "virama plus" sequences, or not?

If a sequence doesn't occur naturally, does a font need to support it?
After all some, Indic (sensu lato) rendering engines deliberately
restrict the number of codepoints of a cluster.

The evidence for some subscript forms is hard to come by.  Some rare
examples of Shan Pali subscript consonants can be found at
https://github.com/notofonts/myanmar/issues/19 .  UTN-11 contains the
self-contradictory statement, "Not every consonant can
be stacked, and while theoretically any consonant can take a subjoined
form, not all implementations will necessarily need to support all
subjoined forms".  Acting on this, the Shan Pali consonants are not
included in the element 'stacked', and Padauk does not support them.

> If not, why was a certain font's lack of support for a "virama plus" 
> sequence mentioned in a derogatory fashion?

The UTS calls out for support of <VIRAMA, WA>, and a one-time
programmer of the font, Martin Hosken, supported assertions of its
existence.

Richard.