On 03/09/2002 11:20:43 AM "Vladimir Ivanov" wrote:

>If I build an Avestan font according to Michael Everson's specifications
>(Avestan characters should be in BMP, see
>http://www.evertype.com/standards/iso10646/pdf/avestan.pdf)

How do you propose to encode the characters? That documentation does not 
specify that, nor could it since actual codepoints are not chosen by 
authors of proposals but by the standards bodies. Did you have in mind 
private-use characters?


> and a Keyman
>keyboard layout, would I be able to type Avestan texts in:
>Word 2002, Publisher 2002, Access 2002 under Windows 2000/XP?

Assuming some answer to the previous question, yes, you can *type* them. 
You'll be lucky to get any more support for them than that.


>Is it necessary to add some special characters to Keyman keyboard to tell
>Uniscribe that Avestan is a
>right-to-left script? 

I suppose you could insert U+202E at the beginning of a run and U+202C at 
the end. Whether it does what you want could only be determined by 
experimentation.


>Or should this be done through the VOLT (Avestan font
>needs some right-to-left kerning)? Generally, how can an application know
>that a certain range of BMP belongs to a right-to-left script?

It has to rely somehow on Unicode character properties. In theory, these 
could be obtained from a variety of sources, including a font (Graphite 
does provide the ability to specify directionality of private-use 
characters, for example), but in practice most software implementations of 
the bidi algorithm don't work that way. Some may have particular ranges 
hard-wired into code; others may lookup in compiled tables.



>Should we wait for Keyman 6 to type Old Persian in the same applications
>because this script is in Plane 1?

Before typing *anything* using Plane 1 codepoints, you should wait until 
it has been incorporated in The Unicode Standard and ISO 10646. If you 
jump the gun and implement something that has been proposed but not yet 
standardised, you create a very real risk of ending up with 
invalidly-encoded data since the codepoints can and often do change 
between a proposal and the approved standard -- even the actual inventory 
of characters often undergoes change.



- Peter


---------------------------------------------------------------------------
Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <[EMAIL PROTECTED]>


Reply via email to