Title:               Comments on Encoding Vedic Accents Proposal

Source:            Unicode Technical Committee

Author:            Lisa Moore, Chair, UTC

Distribution:    Michael Everson, Ireland, WG2

Dr. Om Vikas,  UTC Primary Representative, Ministry of Information             Technology, Government of India

Mike Ksar, Convenor, WG2

Arnold Winkler, Chair, NCITS L2

Action:             For Review and Response  


The UTC considered Mr. Everson’s proposal on Encoding Vedic Accents at the April UTC. During the discussion, a number of questions and concerns were raised.  The UTC believes these concerns should be addressed before it can accept any of the proposed characters into the Unicode Standard.  The questions and concerns were as follows:


Variation Selectors


Ÿ         The vedic accents proposal adds sixteen variation selectors.  The UTC has general architectural concerns with adding the variation selectors and thinks they should be encoded in a common area and handled as a separate proposal.

Ÿ         The UTC is interested in the use of variants, but thinks that variation selectors should not be used to distinguish all glyphs that can represent a character, and they should not be used across annotational traditions.

Ÿ         Does the variation in Vedic forms only apply to Devanagari? What about the usage of variation selectors with other Indic scripts?


Unifications and Other Issues


Ÿ         Regarding the anusvaras and visargas, should the unification be by function or by glyphic appearance? The unification principles to be used are not clear. The exact number of characters proposed is also not clear since the chart might be interpreted as just a list of related glyphic entities in some instances.

Ÿ         Despite the opening text, your document looks very much like an official document.  We encourage you to try to more clearly label it as a proposal.

Ÿ         The UTC is concerned about the combining digits and letters, and is not sure that the separate encoding of combining forms of digits and letters is the best way to proceed.

Ÿ         We should not introduce more combining marks that could be unified with existing combining marks.  If more need to be encoded, they should be encoded as general purpose combining marks, assuming their usage is not clearly restricted to Devanagari.


Encoding Position Issues


Ÿ         The proposed code chart disrupts the phonetic correspondences across Indic scripts. Further consideration should be given in the proposal as to why this is necessary, once it becomes clear exactly how many new characters are required to cover the encoding requirements for Vedic extensions to Devanagari.

Ÿ         Where do you propose encoding these characters - for which plane - the BMP?


We would much appreciate hearing back from you on these questions and concerns in advance of the next UTC, to be held August 8-11.  If you are able to respond, we will most certainly discuss your proposal again at the meeting.


May 31, 2000