ISSUE-61: RCS involving Unicode character categories were found to exhibit interoperability hitch
RCS involving Unicode character categories were found to exhibit interoperability hitch
- State:
- CLOSED
- Product:
- Interoperability Test
- Raised by:
- Takuki Kamiya
- Opened on:
- 2010-11-03
- Description:
- The spec did not specify a specific version of Unicode, which
led some implementations to use version 4.0 while others assumed
the use of version 3.1. We need to find a best way to make this
straight. - Related Actions Items:
- No related actions
- Related emails:
- RE: EXI Telecon Agenda, Wednesday, 10 November 2010 (from john.schneider@agiledelta.com on 2010-11-09)
- RE: EXI Telecon Agenda, Wednesday, 10 November 2010 (from tkamiya@us.fujitsu.com on 2010-11-09)
- RE: EXI Telecon Agenda, Wednesday, 10 November 2010 (from john.schneider@agiledelta.com on 2010-11-09)
- RE: EXI Telecon Agenda, Wednesday, 10 November 2010 (from Youenn.Fablet@crf.canon.fr on 2010-11-09)
- EXI Telecon Agenda, Wednesday, 10 November 2010 (from msc@mitre.org on 2010-11-09)
- ISSUE-61: RCS involving Unicode character categories were found to exhibit interoperability hitch [Interoperability Test] (from sysbot+tracker@w3.org on 2010-11-03)
Related notes:
JS proposed to consider three options.
1. Make the spec assume a specific Unicode version (e.g. 4.0) for
use in computing RCS.
2. Make the spec state the use of a version of Unicode for use in
computing RCS, but still keep it lenient to permit implementations
to use newer version of Unicode. The spec would admonish, however,
that the use of newer versions need to be done very carefully so as
not to face interoperability problems.
3. Firm up the RCS computation rule by stating patterns involving
certain Unicode character categories are *not* subject to RCS
computation.
The WG noted the option #3 appeared to be the best option among the
three, weighing pros and cons of each.
Display change log
