<i18n dev> RFR: 8261621: Delegate Unicode history from JLS to j.l.Character [v2]

[David Holmes]

On 20/02/2021 12:00 am, Naoto Sato wrote:
> On Thu, 18 Feb 2021 14:49:20 GMT, Roger Riggs <rriggs at openjdk.org> wrote:
> 
>>>>
>>>>
>>>> The table is informative and should not be construed as specification.
>>>> The wording "has supported" should be sufficient.
>>>
>>> If this is not specification then doesn't that imply that any provider of any version of OpenJDK would be free to support, or not, whatever version of Unicode that they chose? Surely a minimum supported version must be part of the platform specification?
>>
>> The current version of Unicode is specified in a normative statement just before the table.
>> "Character information is based on the Unicode Standard, version 13.0."
>>
>> The table is not a specification of past revisions.
> 
> RIght. And each Java SE release's spec has the same sentence with the version replaced. In fact, vendors cannot incorporate arbitrary Unicode versions as it would involve API changes. If they wanted to do so, an MR should have to be released.

Roger and Naoto, thanks for clarifying. I took Roger's original comment 
out of context.

David

> -------------
> 
> PR: https://git.openjdk.java.net/jdk/pull/2538
>