Unicode Vs the Rest. Getting the Encoding Balance Right. A Solaris Locale Perspective
Intended Audience: |
Managers, Software Engineers, Content Developers |
Session Level: |
Intermediate |
Statement of Purpose:
Unicode provides support for many scripts and is often viewed as the
solution to all our encoding problems. The reality for many companies
who wish to move to Unicode is that they still need to support both
legacy and new encodings. The former is driven primarily by codeset
conversion issues and the latter by national standards bodies. Getting
the balance right can be challenging. This paper discusses these issues
from the Solaris locale perspective and outlines a locale strategy for
migrating to Unicode which may benefit others dealing with the same
problem.
Paper Description:
Topics to be covered in this presentation are as follows:
- Current status of Solaris support for encodings
- The Solaris locale model
- Locale Proliferation. Controlling locale offerings in a POSIX
environment
- Ground Rules/rationale for adding support for new encodings.
- Engineers Perspective
- Standards Bodies Perspective
- Customers Perspective
- Procurement Issues
- Building,integrating and maintaining Unicode locales in Solaris.
- Font Strategy. Separate or consolidated fonts?
- Migration Strategy. Migrating existing locales to their UTF-8
equivalents. Set customer expectations ahead of time.
- EOL (End Of Life) Strategy. Is there ever a good time to remove
locales?
Conclusions/Recommendations:
- It is not all or nothing as far a Unicode is concerned.
- Legacy encodings must still be supported.
- Support for new encodings must be seriously considered and only
supported if there is a strong business decision to do so.
- Engineering teams need must streamline internal processes in order to
support complex locale offerings.
- Consolidate font offerings where appropriate.
- Have a clear migration path to Unicode and set customer expectations.
- Be careful how you EOL locales. Set customer expectations ahead of
time.
|