From: verdy_p (verdy_p@wanadoo.fr)
Date: Mon Sep 07 2009 - 18:21:15 CDT
"Shriramana Sharma"
> On 2009-Aug-19 00:22, verdy_p wrote:
> > The discussion is not really requesting a name change, but about how
> > an annotation could precise the confusing meaning of the adopted name
> > (still not official, given that the character is still not officially
> > standardized as is, but fixed for the rest of the balloting process,
> > until it finally gets either standardized or rejected).
>
> Is "annotation" same as alias?
Here the term annotation is realted to any non mandatory informational-only property that does not need to be
present within the main UCD file. It may be also presnet only within the annotations found in the charts (but which
are in fact also in a huge extra properties text file used to generate the charts.
The term "alias" takes a stronger meaning: it would imply that this can be used as an alternate name everywhere a
standard name can be used, so it must not collide with all other standard names or aliases and so it brings its own
stability constraints for the choice of standard names for the rest of the standard and its evolutions or additions
(aliases are effectively taking a "slot"' in the same namespace, when an annotation lives in a separate space and
does not have to be unique).
The "annotation" is then simpler to add and without consequences for the standard, and easy to make it visible in
the Unicode charts; but if a good name is judged pertinent, accurate, and reasonnably stable to avoid future
collisions with other characters, why not promoting it later into an alias?
An annotation can then be used temporarily, at least for a long enough experimentation period, if it helps final
users of the script and implementers of renderers, fonts, or text input technologies fixing their usage and
determining a stable and interoperable policy. the nnotation can be modified, amended, fully replaced or dropped at
any time if therre is still some confusion. The same should not happen for aliases that should be stable once they
are defined.
This archive was generated by hypermail 2.1.5 : Mon Sep 07 2009 - 18:24:42 CDT