Fw: Latin Script Danda

[Richard Wordingham via Unicode]

Begin forwarded message:

Date: Fri, 19 Apr 2019 11:30:32 +0100
From: Richard Wordingham <richard.wordingham at ntlworld.com>
To: Shriramana Sharma <samjnaa at gmail.com>
Subject: Re: Latin Script Danda


On Fri, 19 Apr 2019 11:33:35 +0530
Shriramana Sharma via Unicode <unicode at unicode.org> wrote:

> We are using the pipe character as it is readily available in our
> favourite Latin script fonts. See for example:
> https://twitter.com/ShriramanaS/status/793480884116529152  

The broken bar glyph of the pipe character does not feel appropriate.  

> It would be ideal for Sanskrit/Indic text in IAST/ISO to be
> displayable/printable using any common Latin font which is found
> typographically pleasant. For instance the font I have used in that
> Twitter post is Gentium Basic. I use this font for most of my Latin
> script publication purposes (including Unicode documents) and it
> contains the pipe character but it does not contain Devanagari
> characters.  

> It would be difficult to canvas Latin font vendors to include the
> Devanagari characters 0964/0965 on a small technicality of character
> property.  

Font designers for many Indic scripts have had to learn that
U+0964 and U+0965 have the script property of 'Common', not
Devanagari.  I don't trust automated font pickers in that respect,
though.

> Is there a particular reason it's *really* necessary to include Latn
> in the script extension property of 0964/0965?  

No more so than including Indian scripts in the list.  There has been a
threat to use the script extension property in breaking text into
script runs, and U+0964 and U+0965 are often better with
script-sensitive forms.

Richard.