Hi Nandu, 

I will attempt to "pose the question" as best I can. 
(although several very good responses have been given already.) 
Question 1B is pretty well answered.  The others seem to be debatable. 

Q1:  Are there prescribed encoding orders for vedic tone marks 
     relative to 
	A) nasalizations (anusvara, visarga, candribindu) 
     and 
	B) vowel marks (matra)? 

Q2: what is the preferred graphical position of vedic tone marks 
     relative to these same other sets of marks? 

 ---------------------------------------- 

I should point out there are three distinct "orders" of letters and marks: 

1) the phonological, or "logical" order 

2) the order the letters are input or typed 

3) the encoding order, in which they should appear in Unicode text 

4) the order they should or may be displayed (but often a mark is 
displayed over or under another, position isn't really an "order".) 

Different replies on this thread have addressed different ones of these. 

 ---------------------------------------- 

Since I first started asking these questions, I've revisited the 
relevant Unicode and OpenType documents.  And it's clear why this is 
problematic.  The current documents don't quite answer the questions. 

* Unicode Book Chapter 4 Character Properties 
* Unicode Book Chapter 9 South Asian Scripts I 
http://www.unicode.org/versions/Unicode6.3.0/ 
* OpenType terms - especially the "Notation" section 
http://www.microsoft.com/typography/otfntdev/indicot/terms.htm 
* OpenType on Indic Shaping Engine 
http://www.microsoft.com/typography/otfntdev/indicot/shaping.htm 

The Unicde and OpenType standards are pretty clear regarding the 
*encoding* order of tone marks with respect to vowels and consonants. 

OpenType clearly shows 
	[VM] + [SM] (vowel marks before stress/tone marks) 
(And I now realize: the inherant 'a' of each consonant makes it 
consistent to attach these directly to a consonant glyph.) 

 ---------------------------------------- 
However, under any interpretation of the docs I have been able to imagine, 
I do not see anywhere an explicit statement of where nasalizations 
(bindus) such as anusvara and candrabindu may appear in the encoding, 
particularly, where they may be encoded with respect to tone marks. 

There is one rough statement in the Unicode doc 
	"These marks indicate nasalization or final nasal closure of a 
	syllable" 
But does that mean, their encoding must antecede all other marks 
in the syllable? 

Unicode classifies most of the Indic marks with the interesting 
word "reorderant", which I gather to mean their graphcal representation 
might not follow the encoding... I'm really not sure what it means. 

Am I missing something? 

Steve 

AND A FOLLOW-UP 